text
stringlengths 87
880k
| pmid
stringlengths 1
8
| accession_id
stringlengths 9
10
| license
stringclasses 2
values | last_updated
stringlengths 19
19
| retracted
stringclasses 2
values | citation
stringlengths 22
94
| decoded_as
stringclasses 2
values | journal
stringlengths 3
48
| year
int32 1.95k
2.02k
| doi
stringlengths 3
61
| oa_subset
stringclasses 1
value |
|---|---|---|---|---|---|---|---|---|---|---|---|
==== Front
World J Surg OncolWorld Journal of Surgical Oncology1477-7819BioMed Central London 1477-7819-3-711628007410.1186/1477-7819-3-71ResearchPostoperative mortality after cancer surgery in octogenarians and nonagenarians: results from a series of 5,390 patients Damhuis Ronald AM [email protected] Claudia JC [email protected] Willem S [email protected] Rotterdam Cancer Registry, Rotterdam, The Netherlands2 Medical Center Rijnmond-Zuid, Department of Surgery, P.O. Box 9119, 3007 AC Rotterdam, The Netherlands2005 9 11 2005 3 71 71 19 8 2005 9 11 2005 Copyright © 2005 Damhuis et al; licensee BioMed Central Ltd.2005Damhuis et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
To support decisions about surgical treatment of elderly patients with cancer, population-based estimates of postoperative mortality (POM) rates are required.
Methods
Electronic records from the Rotterdam Cancer Registry were retrieved for octogenarians and nonagenarians who underwent resection in the period 1987–2000. POM was defined as death within 30 days of resection and both elective and emergency operations were included.
Results
In a series of 5.390 operated patients aged 80 years and older, POM rates were 0.5% for breast cancer, 1.7% for endometrial cancer and 4.2% for renal cancer. For patients with colorectal cancer, POM increased from 8% for the age group 80–84 to 13% for those 85–89 to 20% in nonagenarians. For stomach cancer, the respective figures were 11%, 20% and 44%.
Conclusion
These results show that resections can be performed at acceptable risk in selected elderly patients with cancer.
==== Body
Background
As a result of ageing of the general population, the proportion of elderly patients with cancer is increasing within Europe. In the Netherlands, 14% of male patients and 17% of female patients are 80 years or older at diagnosis [1]. In the elderly, relative survival is generally worse than in younger patients [2], mainly due to a more advanced stage at diagnosis or due to less extensive treatment. Especially surgery is withheld out of concern for postoperative morbidity and mortality. Many studies, however, suggest that surgical treatment can be performed at acceptable risk and with good results. These reports usually come from specialized centres with selected series and may be too optimistic due to reporting and publication bias. To support decisions about the option of surgical treatment in a general situation, we studied postoperative mortality rates using data from a population-based cancer registry.
Methods
Information on octogenarians and nonagenarians who underwent resection for cancer in the period 1987–2000 was retrieved from the Rotterdam Cancer Registry. The Rotterdam Cancer Registry is operational since 1982 and now covers a region with one university hospital, 15 general hospitals and 2.3 million inhabitants. Trained coding clerks collect information on tumour site and morphology, tumour stage and type of treatment. The clinical record is assessed at least 3 months after diagnosis and notification, thus enabling the evaluation of postoperative mortality, which was defined as death within 30 days of operation. Due to privacy regulations death certificates are not available, thus hampering evaluation of long-term survival.
This study comprises all consecutive patients who underwent primary resection of the tumour, irrespective of the type of surgery, whether being elective, emergency, curative or palliative. To avoid chance findings, the study was restricted to types of cancer with more than 100 patients operated. Subgroup analysis by age was only performed for types of cancer with more than 10 postoperative deaths per subgroup. Postoperative mortality rates were tabulated and differences between subgroups were evaluated with chi-square test.
Results
This study comprises 5,390 patients who underwent resectional surgery between 1987 and 2000. Postoperative mortality increased with age from 5.4% in patients aged 80 to 84 years, to 9.1% in patients aged 85 to 89 years and to 14.4% in nonagenarians (Table 1). In more recent years, a small decrease in postoperative mortality was observed, from 7.7% in the period 1987–1993 to 6.5% in the period 1994–2000 (p = 0.07).
Table 1 Postoperative mortality (POM) in octogenarians and nonagenarians who underwent resectional surgery for cancer
Postoperative mortality
No. of patients No. of deaths %
Sex Male 1385 184 13.3
Female 4005 196 4.9
Age (years) 80–84 3513 190 5.4
85–89 1503 136 9.1
90+ 374 54 14.4
Period 1987–1993 2415 187 7.7
1994–2000 2975 193 6.5
Type of cancer Colorectal 2765 293 10.6
Stomach 424 67 15.8
Breast 1731 9 0.5
Endometrial 350 6 1.7
Kidney 120 5 4.2
Colorectal cancer and breast cancer were the most common types of cancer with postoperative mortality rates of 10.6% and 0.5%, respectively. Postoperative mortality was highest for stomach cancer with 15.8%. For endometrial cancer and kidney cancer, postoperative mortality rates were 1.7% and 4.2%, respectively.
For colorectal cancer, postoperative mortality rates increased from 8% in patients aged 80 to 84 years, to 13% in patients aged 85 to 89 years and to 20% in nonagenarians (p < 0.001) (Table 2). For stomach cancer the respective figures were 11%, 20% and 44% (p < 0.001).
Table 2 Postoperative mortality for octogenarians and nonagenarians with colorectal or stomach cancer.
Type of cancer 80–84 years 85–89 years 90+ years
n POM (%) n POM (%) n POM (%) p-value
Colorectal 1731 8 837 13 197 20 <0.001
Stomach 293 11 99 20 32 44 <0.001
Discussion
Our results show that resectional surgery can be performed at acceptable risk in elderly patients with cancer. It is unknown whether these results can be generalised to other regions, time periods or health care systems. Population-based studies tend to report higher operative mortality rates than studies from single institutions. Selection criteria are presumably different from those in trial series and publication bias hampers honest comparison. Even between population-based studies, selection criteria and definitions may differ. For example, a study from the US [3] analysed operative mortality in patients aged 85 years and older and reported on gastrectomy (16%), colectomy (7%) and nephrectomy (5%), but only included elective operations. After emergency surgery, postoperative death may not be attributable to the resection itself because some patients would also have died after alternative treatment. A second difference is that the US study reported on in-hospital mortality, which produces higher rates than the 30-day definition.
Age-specific reference figures are essential for clinical decision-making but calendar age serves as a poor substitute for biological age, which in itself is difficult to define or determine [4]. Preoperative assessment of the operative risk in geriatric patients is considered very important but current scoring methods provide little assistance [5]. As a consequence, treatment decisions in the elderly cannot be based on general guidelines but will rather require tailored plans that need to be discussed with the patient and his or her family. In some cases, surgery may need to be postponed for the treatment of concomitant disease.
Apart from operative mortality, the remaining quality of life and the life expectancy should be taken into consideration. Non-fatal complications and lengthy hospital stays should not be taken too lightly and may severely impair daily living. The median life expectancy for men and women aged 90 years, is 3.3 and 4.1 years, respectively. The actual life expectancy, however, is difficult to predict and little is known about the course of cancer after suboptimal treatment. Especially in the elderly, preoperative assessment of the extent of disease is crucial to avoid senseless palliative surgery.
For breast cancer, studies have shown that primary hormonal treatment is inferior to direct surgery [6], which is understandable given the low operative risk. For endometrial cancer, the operative risk is less than 2% and alternative treatments should be reserved for patients with severe comorbidity. For kidney cancer, nephrectomy should be considered when urinary function is acceptable, but cancer progression may be slow and a wait-and-see policy is a serious option in patients with smaller lesions. For abdominal surgery, the operative risk is substantial but this risk may still be worthwhile given the absence of other curative options. Whether an operative risk of 44% in nonagenarians with stomach cancer is ethically acceptable, is up to discussion. However, when faced with a guarantee of progressive cancer and no alternatives for cure, patients are willing to accept extremely high-risks [7] that may even seem unacceptable to their physician. For abdominal tumours, emergency presentation is more common in the elderly and obstruction may lead to a quick death in case of non-surgical treatment. Laparoscopic interventions may, however, avert major surgery.
Conclusion
In the near future, the number of elderly patients with cancer will continue to rise due to the ageing of the general population. Judging from the current findings, surgery should not be withheld because of the postoperative mortality but suboptimal or palliative treatment may be necessary in patients with poor physical or mental function. To enable informed decision-making, both patients and clinicians need information on the risks of surgical treatment. This information should be retrieved from population based studies since we cannot rely on the memory of individual clinicians, as is shown by the nine deaths after surgery for breast cancer in a consecutive series of 1731 patients in 15 hospitals over a period of 14 years. Finally, we want to stress the importance of studies incorporating comprehensive geriatric assessment that may provide us with better estimates of the operative risk in the future [8].
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
RD Study management, literature review, preparation of manuscript
CM Study design, statistical analysis
WM Study material contribution, interpretation of data, critical revision of manuscript
Acknowledgements
The Rotterdam Surgical Oncology Working Group for their support.
==== Refs
Visser O Coebergh JWW Dijck van JAAM Siesling S Eds Incidence of cancer in the Netherlands 1998 2002 Utrecht: Association of Comprehensive Cancer Centers
Berrino F Capocaccia R Estève J Gatta G Hakulinen T Micheli A Sant M Verdecchia A Eds Survival of cancer patients in Europe: the EUROCARE-2 study 1999 Lyon: International Agency for Research on Cancer
Finlayson EV Birkmeyer JD Operative mortality with elective surgery in older adults Eff Clin Pract 2001 4 172 177 11525104
Repetto L Venturino A Fratino L Serraino D Troisi G Gianni W Pietropaolo M Geriatric oncology: a clinical approach to the older patient with cancer Eur J Cancer 2003 39 870 880 12706355 10.1016/S0959-8049(03)00062-5
Ramesh HSJ Pope D Gennari R Audisio RA Optimising surgical management of elderly cancer patients World J Surg Oncol 2005 3 17 15788092 10.1186/1477-7819-3-17
Fentiman IS Christiaens MR Paridaens R Van Geel A Rutgers E Berner J de Keizer G Wildiers J Nagadowska M Legrand C Therasse P Treatment of operable breast cancer in the elderly: a randomised clinical trial EORTC 10851 comparing tamoxifen alone with modified radical mastectomy Eur J Cancer 2003 39 309 316 12565982 10.1016/S0959-8049(02)00673-1
Cykert S Risk acceptance and risk aversion: patients' perspectives on lung surgery Thorac Surg Clin 2004 14 287 293 15382760
Audisio RA Bozzetti F Gennari R Jaklitsch MT Koperna T Longo WE Wiggers T Zbar AP The surgical management of elderly cancer patients: recommendations of the SIOG surgical task force Eur J Cancer 2004 40 926 938 15093567 10.1016/j.ejca.2004.01.016
|
16280074
|
PMC1298341
|
CC BY
|
2021-01-04 16:39:03
|
no
|
World J Surg Oncol. 2005 Nov 9; 3:71
|
utf-8
|
World J Surg Oncol
| 2,005 |
10.1186/1477-7819-3-71
|
oa_comm
|
==== Front
World J Surg OncolWorld Journal of Surgical Oncology1477-7819BioMed Central London 1477-7819-3-721628198010.1186/1477-7819-3-72Case ReportRetroperitoneal myolipoma Behranwala Kasim A [email protected] Krissen [email protected] Mona [email protected] Gordon [email protected] Ajay K [email protected] Department of Surgery, Hammersmith Hospital, DuCane road, London, UK2 Department of Pathology, Hammersmith Hospital, DuCane road, London, UK2005 10 11 2005 3 72 72 12 7 2005 10 11 2005 Copyright © 2005 Behranwala et al; licensee BioMed Central Ltd.2005Behranwala et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Myolipoma is a benign tumour in which smooth muscle cells are mixed with adipocytes.
Case presentation
A 34-year old lady presented with a mass in the right iliac fossa detected on computerised tomographic (CT) scan. Wide excision of the retroperitoneal mass was done. Histopathology showed features of myolipoma. There was no recurrence or metastasis at three years.
Conclusion
Myolipoma is a rare benign entity; hence a benign course and good prognosis are expected.
==== Body
Background
Myolipoma is a rare benign tumour, occurring most frequently in adults and is composed of irregularly admixed mature adipose tissue and smooth muscle fibres. Myolipoma have been described in the round ligament [1], spinal cord[2], eyelid[3], subcutaneous[4,5], pericardium[6], retroperitoneum [5,7-9], rectus sheath of the anterior abdominal wall and abdominal cavity with attachment to the abdominal wall[5]. The retroperitoneal tumours described were mostly incidental findings during other operative procedures and seem to be quiescent. [5] It is usually a large tumour with most of the cases reported being at least 9 cm in diameter.
Although the benign nature of this lesion is usually recognized in superficial locations, deeply situated tumours are more likely to be confused with a well-differentiated liposarcoma. [1]
Case presentation
A 34-year old lady presented with intermittent lower abdominal pain associated with abdominal distension for five months. There was no history of gastrointestinal bleeding, nausea or vomiting and the bowel habits were normal. There was no history of seizures, mental retardation, behaviour problems, and skin abnormalities suggestive of tuberous sclerosis. She had undergone a caesarean section, seven months back. On examination there was a midline reducible incisional hernia with minimal tenderness in the right iliac fossa.
Computed tomography scan demonstrated an unexplained right iliac fossa density, which could represent a caecal or pericaecal abnormality (figure 1). An infraumbilical incisional hernia containing bowel loops was also seen. Colonoscopy showed erythema at the ileo-caecal junction. Biopsies from this area were normal. Magnetic resonance imaging confirmed the presence of the mass in the right iliac fossa. It was separate from the caecum and lying inferior to it. The differential diagnosis was an inflammatory mass or a pedunculated lesion from the small bowel such as a pedunculated tumour or even a Meckel's diverticulum. Both ovaries were normal. Operative findings were that of a retroperitoneal mass. Excision of the mass was done with repair of the hernia.
Figure 1 Computed tomography scan showing a mass in the right iliac fossa region.
Macroscopically, the tumour was well circumscribed with a thin fibrous capsule and the consistency was soft. The cut surface showed lobules of fatty tissue with bands and nodules of firm white tissue with a whorled appearance (figure 2).
Figure 2 Macroscopic appearance of the tumour showing lobules of fatty tissue with bands and nodules of firm white tissue.
Microscopic examination (figure 3) showed a lesion composed of lobules of mature fatty tissue harbouring multiple variable sized nodules of smooth muscle fibres. The smooth muscle fibres were arranged in broad interlacing fascicles. No cytological atypia was seen and only one mitotic figure was identified in the multiple sections examined and no necrosis. No proliferation of medium-sized arteries with thick muscular walls was observed. The spindle cells were positive for smooth muscle actin and desmin and negative for CD34 and CD117 (figure 4). The staining for HMB-45 was not performed, as blocks were not available. The features were those of benign myolipoma (lipoleiomyoma).
Figure 3 Nodules of smooth muscle fibres admixed with mature adipose tissue. (Haematoxylin and eosin staining (×40) – Inset ×200).
Figure 4 Immunostaining shows the spindle cells express smooth muscle actin (×40) – Inset (×200).
Discussion
Myolipoma is a rare benign neoplasm, occurring most frequently in adults in the deep soft tissue of the abdomen or retroperitoneum. It is a benign tumour composed of variable amounts of benign smooth muscle fibres and mature adipose tissue with no lipoblasts, floret-like giant cells or zones of atypia.
The term atypical lipoma was proposed for a group of well-differentiated, non-metastasising liposarcomas arising in surgically amenable soft tissues and for deep-seated atypical adipocytic neoplasms that show variation in adipocytic size and atypical stromal cells but lack lipoblasts. However, these neoplasms recur repeatedly and may dedifferentiate and thereby acquire metastatic potential. [10] Atypical lipomatous tumours can undergo dedifferentiation into highly malignant sarcoma [11], a feature not reported in myolipomas.
The main differential diagnoses of myolipoma include well-differentiated liposarcoma, spindle-cell lipoma, angiomyolipoma, leiomyoma with fatty degeneration, lipoleiomyosarcoma, and leiomyosarcomas. Liposarcomas on CT scan show mild enhancement and appear to be poorly marginated and infiltrative. Liposarcomas contain lipoblasts or floret-like giant cells (enlarged eosinophilic cells with atypical nuclei) or zone of atypia. Spindle cell lipoma does not show any positivity for smooth-muscle actin, desmin or vimentin. Two recently described variants of spindle cell/pleomorphic lipoma are also presented: the pseudoangiomatous variant and the dendritic fibromyxolipoma, which corresponds in all likelihood to a peculiar myxoid variant of spindle cell lipoma. [12] Angiomyolipomas contain conspicuous vessels showing thick muscular walls, are HMB-45 positive and frequently associated with tuberous sclerosis. A leiomyoma with fatty degeneration consists of both the components being heterogenously distributed within the mass with the adipose component focally distributed and not as an integral part of the lesion. Leiomyosarcomas of the retroperitoneum invading adipose tissue contain abundance of mitoses in the smooth muscle component and presence of recurrence or metastasis on follow-up.
Tumours consisting of a mixture of mature adipose and smooth muscle tissues, including those designated lipoleiomyomas, fibrolipoleiomyomas and myolipomas, are exceedingly rare. Most often the muscular component is predominant. Soft tissue myolipoma is a benign lesion, which has to be distinguished from lesions with malignant or uncertain biologic behaviour. Pathogenesis of myolipoma remains unclear. There are two main theories, namely adipose metaplasia and a multipotential Mullerian cell origin. [13]
Myolipoma is formed by bundles of spindle-shaped cells with cigar shaped nuclei intermingled with multiloculated clear cells containing small eccentric nuclei. Histologically the tumour is constituted by two components: areas of mature fat tissue intermingled with more cellular areas composed of bundles of spindle shaped eosinophilic cells, reminiscent of smooth muscle cells. In the latter component, cells showing multilobulated, bizarre nuclei are also focally evident. No areas of necrosis or mitosis are found in both the components of the lesion. They are intricate mixtures of adult adipose tissue and bland smooth muscle exhibiting no cellular atypia or nuclear mitotic figures with little vascular proliferation.[5] By immunohistochemistry, the spindle cells express smooth muscle actin and desmin and caldesmone antibodies and contain both oestrogen and progesterone receptors [14]; the clear cells are non-reactive with the immunohistochemical panel, but fat is identified within the cytoplasm. The ultrastructural features of the spindle cells are those of a leiomyoma, while the clear cells are classified as adipocytes. Ultrastructural analysis has shown the coexistence of both adipocytes (lipid containing vacuoles) and smooth muscle cell features (myofilaments) throughout the lesion. [6] The immunohistochemical and ultrastructural findings reveal the myofibroblastic nature of the cellular myoid component of the lesion.
None of the myolipomata reported showed recurrence or metastasis neither did our case. Thus a benign course and good prognosis are expected. Although myolipoma is very rare, pathologists should consider it in the differential diagnosis of fat-containing retroperitoneal masses.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
KB – writing the manuscript
KC – getting the details of the case report and the photographs
MEB – reviewing the slides and providing the photomicrographs
GS – senior Pathologist (confirming the diagnosis) and contributing the pathological aspects of the manuscript
AKK – overall supervision of the writing of the manuscript
Acknowledgements
Written consent was obtained from the patient for publication of his case records and clinical and photomicrographs
==== Refs
Sonobe H Ohtsuki Y Iwata J Furihata M Ido E Hamada I Myolipoma of the round ligament: report of a case with a review of the English literature Virchows Arch 1995 427 455 458 8548133 10.1007/BF00199397
Brown PG Shaver EG Myolipoma in a tethered cord. Case report and review of the literature J Neurosurg 2000 92 214 216 10763695
Sharara N Lee WR Weir C Myolipoma of the eyelid Graefes Arch Clin Exp Ophthalmol 1998 236 630 634 9717661 10.1007/s004170050133
Scarpellini F Pasquinelli G Damiani S Subcutaneous myolipoma with bizarre cells: morphological, immunohistochemical and ultrastructural study of a case and review of the literature Pathologica 1997 89 163 167 9411363
Meis JM Enzinger FM Myolipoma of soft tissue Am J Surg Pathol 1991 15 121 125 1703396
Ben-Izhak O Elmalach I Kerner H Best LA Pericardial myolipoma: a tumour presenting as a mediastinal mass and containing oestrogen receptors Histopathology 1996 29 184 186 8872156 10.1046/j.1365-2559.1996.d01-504.x
Liang EY Cooper JE Lam WW Chung SC Allen PW Metreweli C Case report: myolipoma or liposarcoma – a mistaken identity in the retroperitoneum Clin Radiol 1996 51 295 297 8617045 10.1016/S0009-9260(96)80350-3
Michal M Retroperitoneal myolipoma. A tumour mimicking retroperitoneal angiomyolipoma and liposarcoma with myosarcomatous differentiation Histopathology 1994 25 86 88 7959650
Takahashi Y Imamura T Irie H Tanaka F Fukushima J Fukusato T Harasawa A Shiga J Myolipoma of the retroperitoneum Pathol Int 2004 54 460 463 15144408 10.1111/j.1440-1827.2004.01646.x
Mentzel T Fletcher CD Lipomatous tumours of soft tissues: an update Virchows Arch 1995 427 353 363 8548119 10.1007/BF00199383
Tallini G Erlandson RA Brennan MF Woodruff JM Divergent myosarcomatous differentiation in retroperitoneal sarcoma Am J Surg Pathol 1993 17 546 556 8333554
Guillou L Coindre JM Newly described adipocytic lesions Semin Diagn Pathol 2001 18 238 249 11757863
Dellacha A DiMarco A Foglia G Fulcheri E Lipoleiomyomyoma of the uterus Pathologica 1997 89 737 741 9549382
Fernandez-Aguilar S Saint-Aubain N Dargent JL Fayt I Noel JC Myolipoma of soft tissue: an unusual tumour with expression of estrogen and progesterone receptors. Report of two cases and review of the literature Acta Obstet Gynecol Scand 2002 81 1088 1090 12421182 10.1034/j.1600-0412.2002.811118.x
|
16281980
|
PMC1298342
|
CC BY
|
2021-01-04 16:39:03
|
no
|
World J Surg Oncol. 2005 Nov 10; 3:72
|
utf-8
|
World J Surg Oncol
| 2,005 |
10.1186/1477-7819-3-72
|
oa_comm
|
==== Front
PLoS GenetPLoS GenetpgenplgeplosgenPLoS Genetics1553-73901553-7404Public Library of Science San Francisco, USA 1632788310.1371/journal.pgen.0010061plge-01-05-14Research ArticleA Network of Multi-Tasking Proteins at the DNA Replication Fork Preserves Genome Stability Yeast Genome StabilityBudd Martin E 1Tong Amy Hin Yan 23Polaczek Piotr 1Peng Xiao 1Boone Charles 23Campbell Judith L 1*1 Braun Laboratories, California Institute of Technology, Pasadena, California, United States of America2 Banting and Best Department of Medical Research, University of Toronto, Toronto, Ontario, Canada3 Department of Medical Genetics and Microbiology, University of Toronto, Toronto, Ontario, Canada
Snyder Michael EditorYale University, United States of America* To whom correspondence should be addressed. E-mail: [email protected] 2005 2 12 2005 12 10 2005 1 6 e613 8 2005 12 10 2005 Copyright: © 2005 Budd et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.To elucidate the network that maintains high fidelity genome replication, we have introduced two conditional mutant alleles of DNA2, an essential DNA replication gene, into each of the approximately 4,700 viable yeast deletion mutants and determined the fitness of the double mutants. Fifty-six DNA2-interacting genes were identified. Clustering analysis of genomic synthetic lethality profiles of each of 43 of the DNA2-interacting genes defines a network (consisting of 322 genes and 876 interactions) whose topology provides clues as to how replication proteins coordinate regulation and repair to protect genome integrity. The results also shed new light on the functions of the query gene DNA2, which, despite many years of study, remain controversial, especially its proposed role in Okazaki fragment processing and the nature of its in vivo substrates. Because of the multifunctional nature of virtually all proteins at the replication fork, the meaning of any single genetic interaction is inherently ambiguous. The multiplexing nature of the current studies, however, combined with follow-up supporting experiments, reveals most if not all of the unique pathways requiring Dna2p. These include not only Okazaki fragment processing and DNA repair but also chromatin dynamics.
Synopsis
Maintenance of genome stability from generation to generation is a primary defense against mutation and ensuing disease. Thus, the cell has evolved complex mechanisms, consisting of redundant, partially overlapping pathways, to protect the fidelity of genome inheritance. Using modern genetic screening techniques that allow one to investigate every gene in yeast that might be involved in these pathways, the researchers have defined a network consisting of 322 genes that together safeguard the DNA replication process. Previous approaches were limited to defining the interaction of one or a few genes, but the availability of mutants affecting all of the nonessential yeast genes allowed the identification of over 800 interactions in this study. In addition, the synthetic genetic array technique used in this study allowed identification of every nonessential gene in yeast that interacts with an essential replication protein, Dna2p. The comprehensiveness of the approach identified most, if not all, of the pathways in which the multitasking Dna2p participates, in a single experiment. The genomic scale of the study significantly accelerates understanding of this protein over traditional, low-throughput genetic methods.
Citation:Budd ME, Tong AHY, Polaczek P, Peng X, Boone C, et al. (2005) A network of multi-tasking proteins at the DNA replication fork preserves genome stability. PLoS Genet 1(6): e61.
==== Body
Introduction
In order to preserve the fidelity of genome duplication during DNA replication, cells with complex genomes have evolved a network of pathways composed of the DNA replication apparatus, DNA repair proteins, and regulatory activities. Despite years of general characterization, knowledge of the specific mechanisms by which these pathways are integrated to protect the genome is still incomplete because of the complexity of underlying replication fork processes and their regulation. The challenge in understanding high fidelity genome transmission has progressed from identification and characterization of the individual DNA replication components to investigation of how they combine to form pathways orchestrating repair and regulation.
The first line of defense against genome instability resides with the enzymes of the DNA replication apparatus itself. While the most familiar example is the proofreading activity found in the DNA polymerases, other proteins of the replisome have also evolved substrate specificities to address errors made during replication fork progression. One of these proteins is Dna2p, a helicase/nuclease. The dna2–1 mutation was identified in a screen for yeast mutants defective in DNA replication based on an assay using permeabilized cells [1]. dna2–1 strains were then shown to accumulate subgenomic-size DNA fragments when incubated at the restrictive temperature [2]. Dna2p has both DNA helicase and single-stranded nuclease activities [3,4]. Biochemical and genetic characterization has revealed that Dna2p is involved in the processing of some, but not all, Okazaki fragments. Specifically, it has been proposed that Dna2p acts with FEN1 to remove RNA primers from Okazaki fragments whose 5′ RNA/DNA termini have been extensively displaced by DNA polymerase (pol) δ. Four lines of evidence support a role for Dna2p in Okazaki fragment processing (OFP). First, Dna2p co-purifies with FEN1, which is a structure-specific nuclease required for OFP in the SV40 in vitro replication system [5,6]. Second, overexpression of the Saccharomyces cerevisiae FEN1 gene, RAD27, suppresses the temperature-sensitive (ts) growth defect of a dna2–1 strain, and furthermore, the dna2–1 rad27Δ double mutant is synthetically lethal [7,8]. Third, biochemical reconstitution experiments have shown that excessive strand displacement by pol δ creates long 5′ flaps that are cleaved inefficiently by FEN1, and that initial cleavage of these flaps by Dna2p potentiates more efficient subsequent cleavage by FEN1 [9–13]. Finally, Dna2p prefers to act on flaps with secondary structures in vitro, i.e., hairpins or fold-backs containing CTG repeats, which is probably where helicase functionality becomes necessary [12]. Thus, the Dna2p nuclease has evolved the ideal mechanism for highly specific action at a replication fork, requiring the presence of an unpaired 5′ terminus for activity, and displaying a complete lack of activity for single-stranded gaps in duplex DNA, which would result in recombinogenic double-strand breaks (DSBs). Nevertheless, there is a real question of whether the excessively displaced flaps used to study Dna2p in vitro ever occur in vivo, and some biochemical evidence suggests this may not be the case. Therefore, although biochemical data indicate that Dna2p is required for OFP when FEN1 or pol δ activity is impaired, identification of Dna2p's in vivo substrates requires more appropriate genetic and physiological assays than have been applied to date [14].
The second and third lines of defense for preventing genome instability during DNA replication are the DNA repair pathways and regulatory pathways, such as cell cycle checkpoints. Increasing evidence suggests that these pathways are integrated into the replication pathway through their use of certain replication proteins [15]. Dna2p seems to be one of these multitasking proteins. Besides its function in OFP, our evidence strongly suggests that Dna2p provides a link between DNA replication and DNA repair, since dna2 mutants are sensitive to methyl methane sulfonate (MMS), X-rays, bleomycin, and hydroxyurea (HU) [8,16,17].
To provide a comprehensive view of the roles of Dna2p at the replication fork and in other genome maintenance pathways, we conducted a large-scale synthetic lethal screen by synthetic genetic array (SGA) analysis using DNA2 as a query [18,19]. Two genes are synthetically lethal if single mutants, defective in either gene, are viable, while double mutants, defective in both genes, are inviable. Two mutants are synthetically sick if the double mutant grows significantly slower than either single mutant. Synthetic lethality is useful for identifying redundancy and complementarity, e.g., pathways that compensate for functional deficiencies in each other or genes encoding products that are both required to efficiently process a common substrate, which is often the case for DNA replication and repair proteins. Synthetic lethal screens not only reveal previously unknown genetic interactions with queried mutants but also how gene products and their corresponding pathways functionally associate.
Our results provide a catalogue of most, if not all, pathways that are interdependent with or require Dna2p, thus revealing both the extent and limits of its multitasking character. The work not only confirms a role in OFP, but also identifies functions in (1) a replication/repair helicase subnet, (2) DSB repair and mismatch repair, (3) the replication stress checkpoint, (4) sister chromatid cohesion, (5) chromatin dynamics, (6) histone modification, and (7) osmotic and oxidative stress responses. In a more general sense, the interactions link a specific network of DNA repair and regulatory pathways to a specific network of replication genes that together maintain high fidelity lagging-strand DNA replication.
Results
SGA Screens
Since DNA2 is an essential gene, either a conditional or hypomorphic allele is required for a synthetic lethal screen. We chose two alleles, dna2–1, a ts mutant sensitive to a variety of DNA damaging agents, and dna2–2, a mutant that grows at 23 °C and 37 °C, but is sensitive to MMS, bleomycin, and X-rays [8,20]. dna2–1 contains a P504S substitution in a region of the protein N-terminal to both the nuclease and helicase domains [3]. All enzymatic activities of the Dna2–1 protein are reduced relative to wild-type (WT)—DNA-stimulated ATPase, DNA helicase, and single-stranded DNA nuclease activity [21]. The dna2–2 mutation changes arginine at position 1235, an invariant residue in helicase region IV, to glutamine [8]. A crystal structure of PCR helicase shows that residues in region IV bind the adenine base of ATP [22]. The dna2–1 strain grows very slowly, even at the permissive temperature, so synthetic sickness is sometimes difficult to unambiguously assign for this mutant. The dna2–2 mutant grows faster than the dna2–1 mutant, and synthetic sickness, characterized by slow growth, can be assigned with greater confidence using dna2–2 strains. The use of each dna2 allele in a separate SGA screen expands the range of detectable synthetic lethal interactions.
Table 1 lists the validated synthetic lethal and synthetic sick interactions of dna2–1 and dna2–2 strains obtained from the SGA screen (see Materials and Methods). Validation was performed by preparing a new heterozygous diploid between the respective dna2 allele and the candidate gene, followed by sporulation and tetrad dissection. The dna2-allele specificity of some of the interactions may turn out to be significant, as the mutations differentially affect helicase and nuclease activity. Mutants that appeared to show growth defects in the primary screen but that did not meet the stringent requirements (see Materials and Methods) for interaction imposed by the secondary tetrad analysis are found in Table S1. Mutants synthetically lethal or synthetically sick with either dna2–1 or dna2–2 fall into the following categories: genes involved in OFP (rad27Δ, exo1Δ, yen1Δ, rnh35Δ, rnh202Δ, pol3–01, rpa1, elg1Δ, pol1, and pri1), nonessential helicases involved in maintaining chromosome stability (sgs1Δ, srs2Δ, and rrm3Δ), genes involved in repair (rad52Δ, mre11Δ, rad50Δ, xrs2Δ, sae2Δ, mms1Δ, mms22Δ, slx5, and slx8), genes involved in the DNA replication checkpoint (mrc1Δ, csm3, and tof1Δ), genes involved in chromosomal cohesion (ctf4Δ and ctf18Δ), genes involved in chromatin disassembly/assembly and nucleosome modification and remodeling (spt16, pob3, rad6Δ, bre1Δ, swd1Δ, swd3Δ, hst3Δ, rpd3Δ, pho23Δ, rtf1Δ, and the histone chaperone asf1Δ), genes involved in the oxidative stress response (lys7Δ and sod1Δ), a gene in the osmotic stress response (hog1Δ), and genes involved in degradation of short-lived proteins (ubc4Δ), in polarized cell growth (cla4Δ), and in mRNA processing (trf4Δ and rtt103Δ). Previous studies showed dna2 to be synthetically lethal with several essential genes: mcm10–1, involved in initiation of replication [23]; cdc9, a DNA ligase ([24], although see also [8]); rpa1, a single-stranded DNA binding protein [13]; and spt16 and pob3, two genes involved in chromosome remodeling [8].
Table 1 Genes Identified as Putative DNA2 Interacters from the SGA Screen with dna2–1 and dna2–2 Queries and Verified by Tetrad Dissection
The total number of genes that interact with dna2 to date is 56, a connectivity similar to that found in previous screens [18]. Figure 1 shows two-dimensional hierarchical clustering of the data from the current screen and that of previous screens using 43 genes that interact with DNA2 as queries [18]. Figure 2 provides a graphical representation of the network (consisting of 322 genes and 876 interactions) in which DNA2, RAD27, SGS1, SRS2 (HPR5), RRM3, and POL32 form the six major hubs (largest number of connections). We were surprised that none of these genes showed interactions with cell cycle genes. ctf4Δ, however, which is synthetically lethal with all six of these hub genes and many of the other nodes as well (Figure 2), is synthetically lethal with clb2Δ, clb3Δ, clb5Δ, mad1Δ, mad2Δ, mad3Δ, bub1Δ, bub2Δ, and bub3Δ [18]. Thus, CTF4 may provide a missing link between hub genes and cell cycle and kinetochore functions. The same information is shown in tabular form in Table S2, which contains the interacting genes organized by functional categories defined in the MIPS database (http://mips.gsf.de/genre/proj/yeast/index.jsp). Mutations (identified in studies reported below) that suppress dna2 alleles are also included. We propose that the genes showing the greatest number of interactions encode products that share at least one substrate or have at least one overlapping function.
Figure 1 Clustering Analysis
Hierarchical two-dimensional clustering analysis was applied to the DNA2 interactions and those of 43 genes synthetically lethal with DNA2 (see Table 1). The interactions were clustered with respect to the results of the current screen and 43 previous screens using these genes as queries. A total of 322 genes and 876 interactions, each indicated by a red box, were identified. This panel shows a zoom into the region of most significant overlap of shared genetic interactions. In addition to including reduced fitness genes, this analysis includes genes that give increased fitness with dna2, such as pol32Δ.
Figure 2 A Genome Stability Network
Data were compiled with the Payek program using the dna2 screen results in Table 1 and the results of the previous screens [18,19,92], as well as data compiled for candidate genes synthetically lethal with rrm3Δ [47,75]. The data are presented in tabular form in Table S2.
Okazaki Fragment Processing
Although Dna2p is commonly considered to be required for OFP, this role is far from established. We were therefore interested to find that the SGA experiments identified interactions between DNA2 and additional structure-specific nucleases that, like FEN1, are thought to be involved in overlapping OFP pathways. First, dna2–1 was found to be synthetically lethal with mutations in genes encoding two subunits, rnh35Δ and rnh202Δ, of the main RNaseH in yeast [25], suggesting that Dna2p may be involved in an OFP pathway redundant with RNaseH2.
Second, dna2–2 was synthetically lethal with mutations in genes encoding two exo-endonucleases structurally related to RAD27, exo1Δ and yen1Δ. EXO1 is thought to provide backup function for FEN1 in OFP, since rad27Δ exo1Δ is synthetically lethal (in some genetic backgrounds) and since overexpression of EXO1 suppresses the ts growth of rad27Δ mutants [26–28]. Overexpression of EXO1 also suppressed the temperature sensitivity of dna2–1 at the restrictive temperature of 30 °C (Figure 3), though not at 37 °C [21]. Exo1p nuclease acts on 5′ flap-containing structures, and these genetic interactions suggest that such flaps may be the common in vivo substrate of both Dna2p and Exo1p [29].
Figure 3 Suppression of dna2 by ExoIp Overexpression
Strain W303 dna2–1 was transformed with the plasmids pRS424G and pRS424G-EXO1, an empty plasmid vector and a plasmid expressing ExoIp from the GAL110 promoter, respectively. W303RAD5 strains dna2–1 7A(pRS424G) and dna2–1(pRS424G-EXO1) were grown to mid log phase and serially diluted on yeast-dextrose galactose- and raffinose-containing plates and incubated at 30 °C for 5 d.
We also found that dna2–2 is synthetically sick with yen1Δ at 30 °C and lethal, i.e., ts, at 37 °C. Yen1p shows 23% identity with Rad27p, 33% identity with Rad2p (founding member of this nuclease family and involved in nucleotide excision repair), and 24% identity with Din7p (Rad2p-like endonuclease proposed to be involved in mitochondrial mismatch repair). yen1Δ mutants do not appear to have DNA damage or growth defects, and a yen1Δ exo1Δ din7Δ rad2Δ mutant has no growth defect at either 30 °C or 37 °C [26]. Synthetic lethality with dna2 is the first informative phenotype reported for yen1Δ, although yen1 is synthetically lethal with another gene that is in turn synthetically lethal with several replication mutants (see Protocol S1).
Elg1p, like Dna2p, is proposed to be involved in OFP, as well as in telomere silencing and length regulation [17,30,31]. elg1Δ dna2–1 mutants are synthetically lethal (Table 1). Elg1p is homologous to Rfc1p (the large subunit of the replication factor C clamp loader), to Rad24p (a Rfc1p homolog required for DNA damage checkpoints), and to Ctf18p (another Rfc1p homolog involved in chromosomal cohesion). dna2–2 is also synthetically lethal with ctf18Δ [8], but not with rad24Δ (this work). S phase progression in elg1Δ mutants is slower than in WT strains, suggesting that Elg1p is involved in replication fork translocation [30–32].
Interaction of dna2 with Pol δ
It has been proposed, based on in vitro biochemical reconstitution in vitro, that not all Okazaki fragments require Dna2p for processing. Instead, the in vivo substrates of Dna2p are only those Okazaki fragments on which pol δ produces 5′ flaps longer than 30 nucleotides [10,12,14,33]. We wished to devise a genetic test of this hypothesis. One way to do so is by generating excessive strand displacement in vivo and assessing whether there is then an increased requirement for DNA2. This was accomplished by testing the viability of a strain containing both the dna2–1 mutation and pol3–01, a mutation in pol δ known to increase strand displacement in vitro [33]. As shown in Figure 4, dna2–1 pol3–01 is synthetically lethal. We propose that excessive strand displacement in vivo in the pol3–01 dna2–1 strain causes the lethality, supporting the predictions from purely biochemical reconstitution.
Figure 4 Synthetic Lethality Between dna2 and pol3–01
Strain W303 dna2–1 carrying a TRP1 CEN3 DNA2 plasmid was transformed with an integrating URA3 pol3–01 plasmid [120] cut with BamH1. The transformants were streaked on 5-FOA medium to excise the WT POL3 gene and identify clones with the pol3–01 mutation. Three transformants carrying the pol3–01 mutation (isolates 2, 8, and 11) were restreaked on YPD medium containing 2-amino-5-fluorobenzoic acid (FAA), which selects for strains that have lost the DNA2 TRP1–containing plasmid [121]. Three dna2–1 pol3–01 colonies showing some residual growth on the FAA plates were restreaked on YPD-containing medium. The same three isolates of dna2–1 pol3–01 but containing the DNA2 TRP plasmid, i.e., that had not been grown in the presence of FAA, are shown as controls, as indicated.
Another pol δ subunit mutant, pol32Δ, is synthetically lethal with rad27Δ, and the network of synthetic lethal interactions of pol32Δ is similar to that of rad27Δ, rrm3Δ, sgs1Δ, and srs2Δ, which are also synthetically lethal with dna2 (See Figure 2). We were surprised that pol32Δ did not show up in our SGA screen, and directly tested dna2–1 pol32Δ for synthetic lethality. Rather than showing synthetic lethality, pol32Δ suppressed the slow growth phenotype of the dna2–1 strain at 23 °C and suppressed the ts growth phenotype of the dna2–1 strain at 30 °C (Figure 5A). The pol32Δ mutation did not suppress the lethality of the dna2–1 strain at 37 °C (Figure 5A) or the lethality of a DNA2 deletion (not shown). The pol32Δ deletion mutation also suppressed the DNA damage sensitivity of the dna2–1 and dna2–2 mutants (Figure 5B and 5C, respectively). pol32Δ mutants are sensitive to different concentrations of MMS than are dna2 mutants. dna2–1 and dna2–2 mutants are sensitive to 0.005% MMS, while pol32Δ strains are resistant to 0.005% MMS but are sensitive to 0.01% MMS. Both dna2–1 pol32Δ and dna2–2 pol32Δ mutants showed MMS sensitivity similar to pol32Δ strains rather than to dna2–1 or dna2–2 strains (Figure 5B and 5C). Pol32p is a nonessential subunit of the pol δ holoenzyme and is required for full processivity of pol δ [34]. Therefore, a less processive pol δ suppresses the growth and repair defects of both nuclease- and helicase-deficient mutants of dna2, suggesting that Dna2p is acting on flaps arising from excessive strand displacement by pol δ during DNA repair as well as during OFP. The synthetic lethality of dna2 and pol3–01 and the suppression of dna2–1 by pol32Δ provide strong genetic evidence that the Dna2p substrates shown to be optimal in vitro are also substrates in vivo (see Discussion).
Figure 5 Suppression of Slow Growth and MMS Sensitivity of dna2 Mutants by pol32Δ
(A) WT, pol32Δ, dna2–1 pol32Δ, and dna2–1 strains were grown to log phase, serially diluted, and plated on YPD plates and incubated at 23 °C, 30 °C, and 37 °C for 5 d.
(B) WT, pol32Δ, dna2–2, and pol32Δ dna2–2 strains were grown to log phase, serially diluted, and incubated on MMS-containing YPD plates for 3 d at 30 °C.
(C) WT, pol32Δ, dna2–1 pol32Δ, and dna2–1 strains were grown to log phase, serially diluted, and grown on MMS-containing YPD plates for 5 d at 23 °C.
All strains are isogenic with strain 4741 (Table S4). (dna2–1 grows slowly even at 23 °C, and plates at 23 °C were photographed before they were fully grown so that the other strains would not be overgrown.)
The Helicase Network
We have reported previously that dna2–2 is synthetically lethal with sgs1Δ, srs2Δ, and rrm3Δ [35–37], and our SGA screen also detected these genes. More significantly, there is extensive overlap between genes that are synthetically lethal with dna2 and those that are synthetically lethal with sgs1Δ and srs2Δ, as shown in Figures 1 and 2.
To assess the functional relationship between Dna2p and these helicases, we have tried to further determine whether the synthetic lethality between dna2 and the helicase genes shown in Table 2 stems from defects in DNA replication per se or from the aberrant DNA structures that arise during repair of DNA replication errors. Towards this end, we first checked whether the lethality was suppressed by mutations in genes considered necessary for resolving potentially lethal intermediates that form when the original lesions enter the recombination pathway for repair [38]. The results of our current analysis of dna2–2 sgs1Δ, dna2–2 srs2Δ, and dna2–2 rrm3Δ are summarized in Table 2. dna2–2 sgs1Δ and dna2–2 srs2Δ were synthetically lethal and were either inefficiently or efficiently suppressed, respectively, by rad51Δ, as we reported previously ([35,37]; see Discussion). New here is the finding that dna2–2 rrm3Δ lethality was not suppressed by mutations in the recombination pathway (Table 2). This indicates that cell death in dna2–2 rrm3Δ does not result from the accumulation of nonresolvable recombination intermediates but from the formation of early replication intermediates, or possibly blocked DNA replication forks.
Table 2 Genetic Interactions of DNA2 with Other Helicases
Gene pairs encoding three different heterodimeric complexes (MMS4/SLX3, SLX1/4, and SLD5/8), although they are not DNA helicases, are required for viability in the absence of SGS1 and, thus, are part of the helicase network [39]. We find that only one of these complexes, defined by slx5Δ and slx8Δ, shows synthetic lethal interactions with dna2. Since the function of the Slx5/8p complex is unknown, we cannot yet predict the nature of the substrates that might be shared with Dna2p.
DNA Repair
We have previously shown that dna2 and rad52Δ mutants are synthetically sick [16]. rad52Δ also appeared synthetically sick with the dna2 alleles in the SGA screen. We were surprised that other genes required for recombinational repair did not appear synthetically lethal with dna2 in the SGA screen, since they are synthetically lethal with rad27Δ. To insure that we had not missed such genes, dna2–1 rad51Δ and dna2–1 rad55Δ, as well as dna2–2 rad51Δ and dna2-2 rad59Δ, mutants were constructed. Consistent with the global screen, these double mutants did not appear to be synthetically lethal or sick. We also found that dna2–2 rad51Δ rad59Δ triple mutants were viable (not shown). Our results indicate that the Rad51p/Rad59p portion of the Rad52p pathway does not buffer Dna2p function. MMS1 and MMS22 are two additional presumed DNA repair genes that are also required for normal S phase progression and that show synthetic lethality with dna2 mutants in the SGA screen, as previously described [40]. The Mms1p/Mms22p system may correspond to the pathway that results in dna2–2 rad52Δ growth defects.
Previously, we reported that dna2–2 and rad50–5 double mutants were viable and epistatic for repair [16]. Therefore, one of our most unexpected findings was that dna2 mre11Δ and dna2 rad50Δ strains are synthetically lethal (see Table 1). rad50–5 is a point mutation of RAD50 that is as sensitive to irradiation as either a rad50Δ or a rad52Δ strain. Clearly, some function must remain in the point mutant as compared to the deletion mutant, however, since the dna2–2 rad50–5 strain is viable while the dna2–2 rad50Δ strain is inviable. MRE11, RAD50, and XRS2 encode members of the Mre11p complex, which is required for the intra-S phase checkpoint, for homologous recombination, for non-homologous end joining, and for telomere maintenance ([41,42] and references therein). Although xrs2Δ was not found in our SGA screen, we have since shown that dna2–2 xrs2Δ is synthetically lethal (this work).
dna2 Mutants Do Not Require the rad9, mrc1, mec1, or tel1 Checkpoint Activities for Viability
Our analysis of the interaction between dna2 and checkpoint mutants supports the view that dna2–1 and dna2–2 mutants accumulate less damage than rad27Δ mutants. First, we found that neither dna2 allele used in our screen is synthetically lethal with rad9, a mediator of the intra-S DNA damage checkpoint, as found previously [8,43] for other dna2 alleles. rad27Δ mutants are synthetically lethal with rad9Δ. In addition, rad27Δ mutants are synthetically lethal with the DNA damage checkpoint mutants rad24Δ and rad17Δ, which represent the checkpoint clamp-loader-like complex and the checkpoint clamp-like proteins, respectively, but dna2 rad24Δ and dna2 rad17Δ are both viable (this study). Thus, dna2 mutants do not accumulate sufficient damage to require the intra-S DNA damage checkpoint for viability at the permissive temperature.
We found in the SGA experiments, however, that dna2 mutants are synthetically lethal with mrc1Δ, tof1Δ, or csm3Δ (Table 1). These genes define a checkpoint pathway thought to operate in parallel with the Rad9p pathway to respond specifically to replication stress. Mrc1p, Tof1p, and Csm3p are required for Rad53p phosphorylation in the presence of HU or MMS, both of which induce S phase DNA damage [18]. Mrc1p is an adapter molecule in the checkpoint, whose phosphorylation by Mec1p at SQ/TQ motifs is required for activation of downstream effectors [44,45]. dna2 mrc1Δ synthetic lethality seemed to contradict the observation that other dna2 alleles actually show improved survival in the absence of the checkpoint ([43]; see below), and suggested that such synthetic lethality might be due to another function of MRC1. MRC1 and TOF1 appear to play a direct role in yeast DNA replication, as well as in the checkpoint. mrc1Δ strains also show a slow S phase, and Mrc1p and Tof1p have been localized to moving replication forks [46]. Osborn and Elledge [45] constructed a separation-of-function mrc1 mutant that has all 17 TQ and SQ Mec1p target sites mutated to non-phosphorylatable AQ. This mutant, mrc1AQ, like mrc1Δ, is checkpoint defective, as evidenced by the fact that Rad53p phosphorylation is blocked in mrc1AQ rad9Δ mutants upon treatment with HU or MMS. However, mrc1AQ mutants are replication proficient. The mrc1AQ mutant allowed us to ask whether the replication defect of mrc1Δ was responsible for dna2 mrc1Δ synthetic lethality. dna2–1 mrc1AQ and dna2–2 mrc1AQ strains were constructed and were viable and did not appear synthetically sick, indicating that the checkpoint function of Mrc1p is not required for viability in dna2 mutant backgrounds (Figure 6). In order to be sure that RAD9 was not substituting for MRC1 in the dna2–2 mrc1AQ mutant, dna2–2 mrc1AQ rad9Δ mutants, which lack both the DNA damage and replication stress checkpoints, were constructed (Figure 6). The viability of dna2–2 mrc1AQ rad9Δ strains suggests that it is inactivation of the replication function of MRC1 that is responsible for synthetic lethality in the dna2 mrc1Δ mutants.
Figure 6 Separation-of-Function Checkpoint Mutants mrc1AQ and rad9Δ Are Not Synthetically Lethal with dna2 or rrm3Δ Mutants
mrc1Δ experiments were carried out in the isogenic 4741 strain and are listed in Table S4. mrc1AQ experiments were carried out in strains isogenic with W303 RAD5
+.
Panel 1 (top): segregants from a DNA2/dna2–2 MRC1/mrc1Δ diploid. Segregants are isogenic with 4741. d, dna2–2; m, mrc1Δ.
Panel 2: segregants from a MRC1/mrc1AQ RAD9/rad9Δ DNA2/dna2–2 diploid. d,
dna2–2; m, mrc1AQ; r, rad9Δ.
Panel 3: segregants from a DNA2/dna2–1 MRC1/mrc1AQ RAD9/rad9Δ strain. d, dna2–1; m, mrc1AQ; r, rad9Δ.
Panel 4 (bottom): segregants from a RRM3/rrm3Δ MRC1/mrc1AQ diploid. m, mrc1AQ; r, rrm3.
mrc1Δ and rrm3Δ mutants are also synthetically lethal [47] and, because RRM3 and DNA2 interact (see Table 2), we asked whether the mrc1Δ rrm3Δ synthetic lethality is caused by a checkpoint or replication defect. The mrc1AQ rrm3Δ double mutants showed the same viability as the single mutants (Figure 6). Since dna2 and rrm3 are also synthetically lethal, this suggests that DNA2, RRM3, and MRC1 functions may be interdependent in DNA replication.
To further test the idea that S phase checkpoint signaling is dispensable for dna2 mutant viability, the interaction of dna2 with mec1Δ and tel1Δ, mutations in genes upstream of MRC1 and RAD9 in the checkpoint, were investigated. A dna2–2/DNA2 tel1Δ/TEL1 mec1Δ/MEC1 sml1Δ/SML1 heterozygote was sporulated, and Table S3 lists the genotypes obtained among the tetrads. (sml1Δ allows for mec1Δ viability.) The mec1Δ mutation partially suppressed the slow growth phenotype of dna2–2 strains. Thus, as previously observed for dna2–20 [43], the Mec1p-mediated checkpoint is deleterious in the dna2–2 mutant (Figure 7). The tel1Δ mutation shows negative synergy but not lethality with dna2–2 at 37 °C (Figure 7). The negative synergy between dna2–2 and tel1Δ is evidence that Dna2p and Tel1p may function together at DSBs and/or at telomeres, along with the Mre11p complex [48]. dna2–2 tel1Δ mec1Δ triple mutants were recovered. As shown in Figure 7, however, the dna2–2 tel1Δ mec1Δ mutant grew more slowly than any of the single or double mutants. The telomere defects of the mec1Δ tel1Δ strain caused it to senesce as rapidly as est2Δ (telomerase catalytic subunit deleted) strains [49]. The dna2–2 mutation may cause additional defects in telomere replication. Thus, enhanced telomeric senescence might account for the slow growth of the dna2–2 tel1Δ mec1Δ mutant, just as we have shown that dna2–2 est2Δ is synthetically lethal due to accelerated senescence [17].
Figure 7 Synthetic Lethality of mec1Δ tel1Δ with dna2–2 Mutations
Strains used in these experiments are listed in Table S4, and were isogenic or congenic with W303 RAD5
+. Segregants of a MEC1/mec1Δ TEL1/tel1Δ DNA2/dna2–2 SML1/sml1Δ diploid were placed on a YPD plate incubated at 30 °C (A) or 37 °C (B).
Finally, we tested the interaction of DNA2 with the checkpoint effector kinase RAD53. We crossed dna2–1 and dna2–2 to an isogenic rad53Δ sml1Δ strain. dna2–1 rad53Δ sml1Δ and dna2–2 rad53Δ sml1Δ mutants were fully viable. The viability of dna2 rad53 strengthens the conclusion that the DNA damage arising in a dna2 mutant is not sufficient to require the S phase checkpoint for viability. (dna2–1 mutants, however, do induce an amount of damage above the threshold for checkpoint activation at restrictive temperatures, since they arrest at the metaphase to anaphase transition in a MEC1-dependent manner [M. E. B. and J. L. C., unpublished data].)
Sister Chromatid Cohesion and Repair of DSBs in the rDNA
Ctf4p is a pol α–binding protein [50], and ctf4Δ strains are defective in sister chromatid cohesion [20]. dna2–2 was identified as a mutant synthetically lethal with ctf4Δ, but we have shown that dna2 mutants are not defective in cohesion [51]. We previously reported that the dna2-2 mutation gave rise to an increased frequency of DSBs at the replication fork barrier (RFB) in the rDNA and that deleting FOB1, which is required for pausing at the RFB, suppressed DSB formation. A reasonable explanation for all of these observations is that the DSB damage sustained by dna2–2 mutants at the RFB might require Ctf4p-mediated sister chromatid cohesion for repair. If so, then one would expect fob1Δ to suppress dna2–2 ctf4Δ synthetic lethality. We dissected 55 tetrads from a dna2–2 ctf4Δ fob1Δ heterozygote and incubated the spores at 30 °C. Viable dna2–2 ctf4Δ fob1Δ mutants were obtained in the expected numbers, demonstrating suppression. Although the dna2 ctf4Δ fob1Δ triple mutants grew at 23 °C and at 30 °C, they did not grow at 37 °C and were highly sensitive to X-rays (Figure 8). The behavior of the triple mutant indicates that defects in the rDNA locus are critical for some of the phenotypes of dna2–2, but that defects elsewhere throughout the chromosome must still occur, giving rise to ts growth and DNA damage sensitivity.
Figure 8 Deletion of FOB1 Suppresses dna2 ctf4Δ Synthetic Lethality
Top panel: Tetrads from the dissection of a DNA2/dna2–2 CTF4/ctf4Δ FOB1/fob1Δ heterozygote. c, ctf4Δ; d, dna2Δ; f, fob1Δ; WT, DNA2 CTF4 FOB1. The following spores were recovered: 11 WT, 21 dna2–2, 26 fob1Δ, 17 ctf4Δ, 18 dna2–2 fob1Δ, 20 ctf4Δ fob1Δ, zero ctf4Δ dna2–2, and six dna2–2 ctf4Δ fob1Δ. Since the triple mutant grew slowly at 30 °C, another 27 spores were dissected and incubated at 23 °C. The following spores were recovered: 14 WT, nine fob1Δ, 14 ctf4Δ, seven dna2–2, 14 dna2–2 fob1Δ, nine ctf4Δ fob1Δ, zero dna2–2 ctf4Δ, and nine dna2–2 ctf4Δ fob1Δ. The triple mutant did not grow at 37 °C.
Bottom panel: X-ray sensitivity of dna2–2 ctf4Δ fob1Δ. Strains with genotype WT, fob1Δ, dna2–2 fob1Δ, ctf4Δ fob1Δ, and dna2–2 ctf4Δ fob1Δ were grown to log phase, irradiated as described in Materials and Methods, serially diluted, plated on YPD plates, and incubated at 30 °C. Strains are isogenic or congenic with 4741.
Nucleosome Remodeling: Dna2p Interacts with Pol α and Primase
Pob3p and Spt16p/Cdc68p form a heterodimer that is a component of the ATP-independent chromatin remodeling activity yFACT [52]. dna2–2 is synthetically lethal with a non-ts allele of POB3, pob3–21 [52], and various alleles of spt16 are synthetically lethal or sick with dna2–2 [53]. Since yFACT may participate in both DNA replication and transcription, to investigate a potential link with the role of Dna2p in replication and/or repair, we took advantage of the observation that yFACT interacts both genetically and physically with pol α [54,55]. This suggested dna2 might be synthetically lethal with a mutant containing a pol α protein that fails to interact with Pob3p/Spt16p, pol1–1 (with glycine at position 493) [56]. We established the synthetic lethality of dna2–2 and pol1–1 (18 tetrads, 46 viable spores, no double dna2–2 pol1–1 mutants). The dna2–2 pol1–1 lethality is allele specific, since dna2–1 is not synthetically lethal with pol1–17, a catalytic site mutant [57] (see Discussion). Although yFACT may also affect transcription and the synthetic effects between dna2 and yFACT components could be due to reduced transcription of dna2–2 or of other replication genes in the double mutants, the synthetic lethality of dna2–2 pol1–1 argues that Dna2p and yFACT may interact during DNA replication.
We went on to investigate genetic interactions between dna2 and genes encoding other pol α subunits. We found that dna2–2 is synthetically lethal with a primase subunit mutant, pri1-M4 (18 tetrads, 51 viable spores, no double mutants). This result further implicates Dna2p in lagging-strand DNA replication, as the pri1-M4 mutant is defective in elongation [58].
Chromatin Remodeling and Histone Modification
In addition to clarifying the relationship between dna2 and the yFACT complex by identifying the dna2/pol1–1 interaction, the SGA screen suggested a role for Dna2p in additional complexes that control assembly and disassembly of chromatin during DNA replication. We found that dna2 mutants are synthetically lethal or synthetically sick with asf1Δ, defective in a chaperone protein involved in histone deposition at the replication fork. Since dna2 mutants are defective in OFP, they might also be defective in the rapid reassembly of chromatin behind the replication fork, leading to an accumulation of defective chromatin and lethal DSBs [59].
The SGA screen also revealed that dna2 mutants are synthetically lethal or sick with rad6Δ, bre1Δ, swd1Δ, and swd3Δ (see Table 1). Rad6p is a ubiquitin-conjugating enzyme (E2) required for post-replication repair, the N-end rule, and chromatin modification [60,61]. Rad6p functions with two alternative ubiquitin ligases (E3)—RAD18 in post-replication repair and BRE1 in histone H2B ubiquitylation [62–65]. dna2 rad18Δ mutants grew normally (this work). By contrast, dna2 showed synthetic sick interactions with bre1Δ and with mutations in two genes that act downstream of BRE1, swd1Δ and swd3Δ (Table 1). Bre1p-mediated histone H2B ubiquitylation is necessary for histone H3 lysine 4 methylation by Swd1p and Swd3p. These results may indicate that dna2 rad6Δ growth defects do not result from defective post-replication repair, but rather from defects in histone modification.
dna2 is also synthetically lethal with rtf1Δ, which is involved in recruitment of Set1p and H2B ubiquitylation. RTF1 is a member of the PAF1p complex. Since the PAF1p complex is involved in transcription, a trivial explanation of the dna2 rtf1Δ lethality could be that transcription of dna2 or of another replication gene is reduced and indirectly creates the synthetic lethality. If so, then dna2 should also be synthetically lethal with a paf1Δ. However, we found that no dna2 allele tested was either synthetically sick or lethal with paf1Δ. Recent evidence suggests that Rtf1p, which recruits Set1p, affects genome-wide ubiquitylation of histone H2B by a mechanism that is distinct from its function as a transcriptional elongation factor [66], and that would explain how we can find no effect of paf1Δ but lethality with rtf1Δ.
dna2 also shows synthetic interactions with mutations in genes encoding histone deacetylases: hst3Δ and pho23Δ. Rpd3p and Pho23p are components of a histone deacetylase complex that is thought to be involved in regulation of initiation of DNA replication [67,68], and therefore we tested dna2–2 rpd3Δ, and found that it was synthetically lethal. DNA2 is the only replication gene that has shown a genetic interaction with rpd3Δ to date (see also Protocol S1). (We wish to emphasize that for all of the chromatin-modifying pathways detected in the screen, we have shown, in studies in preparation for presentation elsewhere, that the effects are not due to the trivial explanation of reduced transcription of the mutant dna2 genes).
Other Interactions
Additional nodes in the network (oxidative stress genes, osmotic stress genes, and genes involved in RNA modification/catabolism [mutants trf4Δ and rtt103Δ]) that are synthetically lethal with dna2 are described and discussed in Protocol S1. Genes involved in transcription elongation (CAF20, for example) have not yet been further analyzed because they may reduce transcription of Dna2p or another replication protein and thus indirectly cause lethality.
Discussion
We find that 56 genes interact genetically with DNA2. Comparison of our results with those of previous synthetic lethal screens using 43 of the DNA2-interacting genes defines a set of pathways, all of which are interdependent with DNA2 and that form a network for preserving genome stability (see Figure 2). The six major hubs shown in Figure 2 link DNA replication, DNA repair, chromosome dynamics, checkpoints, chromosome structure/chromatin, osmotic stress, oxidative stress, and RNA metabolism. A major link to the cell cycle and the kinetochore occurs through a single gene, CTF4. Analysis of mutants that give rise to gross chromosomal rearrangements, the type of damage considered to be the most likely result of replication apparatus failure, identifies the same pathways [69–75]. The comprehensive nature of the SGA screen, however, allows greater insight into the structure of the network that coordinates these events. It is striking that this topology can be superimposed on the prokaryotic DNA replication interactome recently identified using protein–protein interactions [76]. The bacterial genome maintenance network consists of many of the E. coli orthologs (i.e., pol III holoenzyme, SSB, RecQ, RecG, SbcB, and RecJ) of the yeast replicative polymerase, its subunits, and the helicases and nucleases that form hubs and major nodes in our genetic network [76]. The common denominator in the diverse approaches was the use of a replication gene or protein as the bait. The parallels between the organisms point to evolutionary conservation in the coordination of processes that protect the genome. We suggest that the complexity of such processes was required for the evolution of large genomes, where the fidelity of the replication apparatus itself could not guarantee a sufficiently high level of accuracy and stability in genome transmission.
Current methods of scoring interactions do not result in identification, in the initial SGA screen, of every interacting gene (see discussions in [18]). To approach completeness, following identification of a single gene in a pathway, e.g., MRE11, in the SGA screen, we pursued “traditional” investigation, on a gene-by-gene basis, of other genes in the putative pathway, such as XRS2. Similarly, the identification of the Bre1p ubiquitylation pathway was interpreted only after testing downstream genes in the histone H2B modification pathway. This type of comprehensive genetic analysis is a powerful new tool for rapidly characterizing the full complement of processes requiring replication genes that might be coming under analysis for the first time, as well as rendering a coherent picture of years of genetic analysis of other genes. The genetic screen then enables one to rationally design experiments to determine, in molecular terms, the contribution of the replication protein to these processes. The outcome of such secondary analyses of DNA2 is discussed and interpreted below.
Stronger Links between Dna2p and OFP during Lagging-Strand Replication: RNH32, RNH202, EXO1, YEN1, POL3, POL32, POL1, and PRI1
Structure-specific nucleases.
Although it has been known for some time that dna2 and rad27Δ are synthetically lethal, convincing genetic interactions between dna2 and other lagging-strand activities have not been previously identified. As shown by our demonstration in this work of synthetic lethality of dna2 and rtf1Δ and the viability of dna2 and paf1Δ (see Results), synthetic lethality with one gene in a pathway does not prove interaction with other genes in that pathway. Therefore, the identification of so many additional lagging-strand genes in the current study is a matter of some note. The synthetic lethality of dna2 with the genes encoding structure-specific nucleases (RNH32, RNH202, EXO1, and YEN1) provides the first evidence, to our knowledge, that RNA may be a substrate of Dna2p in vivo and strengthens evidence, as predicted from our in vitro studies [14], that Dna2p acts primarily if not exclusively on flaps in vivo. It has been known for some time that rad27Δ rnh35Δ is synthetically sick, but not lethal [77]. Since FEN1 is generally considered the major OFP nuclease, the more significant synthetic lethality of dna2 rnh35Δ and dna2 rnh202Δ was somewhat unexpected, and suggests that Dna2p also acts in an OFP pathway redundant with RNaseH2. Since Rnh35p does not act on flap-containing structures, the common substrate for RNaseH2 and Dna2p is probably RNA. Alternative to a role in OFP, the dna2 rnh35Δ synthetic lethality might reflect a redundant role for RNaseH2 and Dna2p in mRNA processing, an additional function proposed for RNaseH2 [78]. rnh35Δ mutants also have very short telomeres [79]. Since Dna2p is required for de novo telomere synthesis and dna2 est1 and dna2–2 est2 are synthetically lethal [17], dna2–1 rnh35Δ lethality could be related to events at telomeres.
Suppression of dna2 by EXO1 overexpression, combined with dna2 exo1Δ synthetic lethality, suggests that Exo1p can provide a backup for Dna2p in OFP, with 5′ flaps constituting the common substrate. exo1Δ is also synthetically lethal with mre11Δ, and it has been proposed that Exo1p participates in the resection of ends at DSBs in preparation for recombinational repair [42]. Exo1p may also be involved in a late step in the Msh2p-dependent mismatch repair pathway and perhaps in telomere end processing [27], which might also account for the dna2 exo1Δ lethality. There is no biochemical evidence as yet to show whether Yen1p is a structure-specific 5′ to 3′ nuclease, but this is likely given its similarity to FEN1 and Exo1p. Given that Yen1p is a preferential substrate of the Clb5p cyclin-dependent kinase required for proteins that function in G1 and early S phase [80], the yen1Δ dna2–2 synthetic interaction may reflect a direct link between YEN1 and DNA replication (see also Protocol S1). Since there are about 100,000 Okazaki fragments (1.5 × 107 bp per genome/100 bp per Okazaki fragment) during S phase, it is not surprising that multiple nucleases are involved in OFP.
Pol δ.
Our genetic analysis supports the notion that Dna2p processes a specific subset of Okazaki fragment flaps—namely, those arising from excessive strand displacement by pol δ. We showed not only that dna2 is synthetically lethal with a mutant of pol δ that gives abnormally high levels of strand displacement, but also that dna2 is suppressed by a mutant that gives decreased strand displacement. pol3–01 is a mutation that inactivates the 3′ to 5′ exonuclease function of pol δ. This nuclease has three known functions [81]. First, because pol3–01 is a strong mutator, we can surmise that one function is in proofreading during DNA polymerization. Second, negative synergy of pol3–01 with msh2Δ and exo1Δ suggests a role in mismatch repair. These two activities are probably not relevant to the dna2 pol3–01 synthetic lethality since dna2–1 is not a mutator mutation and thus catastrophic mutagenesis is not likely to occur in the dna2 pol3–01 mutant ([8]; unpublished data). OFP is the third process proposed to require pol δ exonuclease, since pol3–01 is synthetically lethal with rad27Δ [81]. Mechanistically, this has been rationalized by the fact that pol δ 3′ exonuclease is an inhibitor of the strand displacement activity of pol δ during in vitro DNA synthesis [33]. In the absence of its 3′ exonuclease, pol δ can no longer idle at a nick, since the 3′ displaced flaps that form as intermediates in idling cannot be removed nucleolytically. Instead, pol δ processively displaces the strand having a 5′ terminus [33]. This is consistent with the idea of dna2 pol3–01 lethality stemming from the failure to efficiently process excessively long 5′ flaps on Okazaki fragments. pol δ-DV is another exonuclease-deficient allele of pol δ, and rad27-d is a proliferating cell nuclease antigen noninteracting mutant of rad27. Overproduction of Dna2p suppresses the lethality of a pol3 d-DV rad27-p rad51Δ strain [11], supporting our conclusions.
dna2–1 and dna2–2 mutants are suppressed by deleting the nonessential POL32-encoded subunit of pol δ that is required for optimum strand displacement. Pol32p is required for efficient in vitro DNA replication by pol δ in the presence of replication factor C, proliferating cell nuclease antigen, and a primed template. The ability to displace 5′ ends is drastically decreased for pol δ lacking Pol32p [35,43], and the pol δ complex is expected to be defective in strand displacement synthesis in pol32Δ strains, thus reducing the need for Dna2p. Pol32p has also been shown to interact with pol α, and the same mutant that is synthetically lethal with dna2–2, pol1–1 (G493R), is synthetically sick with pol32Δ [82]. This might hint at coordination between Okazaki fragment initiation and elongation, although there is no reported phenotype associated with a mutation (pol32–8) that disrupts the Pol32p–pol α interaction. While this manuscript was being prepared, it was reported that a mutation in cdc27
+, the Schizosaccharomyces pombe ortholog of POL32, suppresses one allele of S. pombe dna2 [83], so this suppression is conserved. Why is pol32Δ rad27Δ lethal [18] while pol32Δ dna2 grows more robustly than dna2 mutants? Loss of pol32Δ may shift the course of OFP from a flap removal pathway to one employing RNaseH. In the RNaseH pathway, FEN1 exonuclease may become essential to remove the last ribonucleotide, an activity that Dna2p does not appear to possess.
Interaction with Pol α and Primase (and Mcm10p).
Our demonstration here of synthetic lethal interactions of dna2 with pol1–1 and with pri1-M4, components of pol α–primase, may also fortify the argument that Dna2p participates in OFP. Although pri1-m4 has an S phase checkpoint defect in addition to a DNA replication defect, the replication defect is more likely to be responsible for the synthetic lethality with dna2, given our data that dna2 is not synthetically lethal with any of the mutations in major checkpoint genes, including mec1Δ sml1Δ. Certain alleles of dna2 are also synthetically lethal with mcm10–1 [23]. Recently, MCM10 has been implicated in elongation and in stabilizing pol α in vivo as well as in stimulating pol α in vitro [84,85]. The interdependent functions of DNA2 and MCM10 may reflect an interaction in lagging-strand replication. Alternatively, Dna2p might play a role in repair of mcm10–1-generated damage. The combined new data on interactions between Dna2p, pol Δ, pol α, and primase may be evidence for a previously unexpected coupling of primer synthesis, polymerase switching, and primer removal.
Differences between the Genetic Interactions of DNA2 and Those of RAD27
Comparison of the data presented here (see Figure 2; Table S2) and the results of a similar thoroughly validated SGA screen using rad27Δ as a query gene [86] reveals a wide (and unanticipated) divergence between genes that are synthetically lethal with dna2 and those that are synthetically lethal with rad27Δ. This divergence implies that the two enzymes may have slightly different sets of substrates. As pointed out in Results, the synthetic lethality of rad27Δ [69,87], but not dna2, with checkpoint mutants and with recombination mutants suggests that rad27Δ mutants probably accumulate more single-stranded DNA (the proposed signal for checkpoint activation) and more DSBs (repaired by recombination) [87], than dna2 mutants do. If dna2 mutants accumulate less damage than rad27Δ mutants, this in turn might suggest that FEN1 is the major OFP nuclease and that Dna2p is required at fewer (or different) sites than FEN1 [88]. This conclusion is consistent with other previously published evidence (despite the potential conundrum that deleting DNA2 is lethal and deleting RAD27 is not). First, pol δ, proliferating cell nuclease antigen, and FEN1 appear to act in a highly concerted fashion on templates that are optimal for pol δ efficiency in vitro, with little evidence that flaps longer than a few nucleotides are ever produced [10,33]. Second, dna2 mutants are weak mutators, while rad27Δ mutants are strong mutators, as measured by point mutations or stability of di- or trinucleotide repeats, or even larger repeats [16,87,89–91]. Dna2p may be specialized to function in OFP in genomic locations where the DNA sequence poses problems for pol δ, creating flaps that are not good substrates for FEN1. These regions are likely to include the rDNA and telomeres, since we have shown significant replication defects in these loci in dna2 mutants [17,35–37]. The role of Dna2p is not likely to be limited to these regions, however, as our previous immunofluorescence and chromatin immunoprecipitation analyses show Dna2p to be located at many other genomic regions during S phase [17]. Possible sites are replication slow zones or the inverted Ty repeats that give rise to genomic instability.
If FEN1 is the major flap nuclease, either Dna2p might help FEN1 on some flaps or Dna2p might recognize discrete subsets of flaps and process them independently. The genetic differences observed, combined with the quantitative biochemical data that show that Dna2p is very inefficient at stimulating FEN1, even on long flaps, direct future attention to potentially independent roles for Dna2p and FEN1.
The Helicase Network for Preserving Genomic Stability
Previous screening of the nonessential gene knock-out collection with sgs1Δ and srs2Δ as queries identified a so-called helicase network defined by a set of common interactions [92]. Genes in this network are implicated by dozens of recent studies in the repair of damaged replication forks through sister chromatid recombination and replication restart mechanisms, but coupling of the network to DNA replication remains poorly understood (e.g., [93]). By reversing the screening process and using an essential lagging-strand replication gene as query, we have found that dna2 is synthetically lethal with mutations in all of these helicases and in the genes with which they interact. The interactions shown in Figures 1 and 2 and our subsequent work (see Table 2) suggest that Dna2p may be one of the major replication proteins that coordinate this helicase network and replication.
The synthetic lethality of dna2–2 sgs1Δ and its lack of suppression by rad51Δ (Table 2) suggests that Sgs1p participates directly in DNA replication by aiding Dna2p in stimulating flap cleavage during OFP under some circumstances. This interpretation is attractive since the human Sgs1p orthologs, BLM and WRN, interact physically with Dna2p and suppress the replication defects in dna2–1 mutants when overexpressed in yeast [94,95]. (The suppression could not be investigated with yeast SGS1, since its overproduction is toxic [A. Morgan, personal communication; unpublished data].) The reproducible effect of deleting RAD51 in restoring defective growth to dna2–2 sgs1Δ mutants, however, is consistent with SGS1 playing an additional role in a late stage of recombinational repair of dna2-induced lesions, such as in the resolution of Holliday junction intermediates into viable products, a reaction that requires Sgs1p in conjunction with Top3p [96]. Dna2p might be required to remove 5′ ends of nascent DNA in reversed forks, while Sgs1p serves as a helicase to resolve the forks, like RecJ and RecQ in bacteria [97]. Lesions due to DNA2 deficiency do have the potential to lead to lethal recombination intermediates, because we find that dna2 is synthetically lethal with srs2Δ and that this lethality is efficiently suppressed by deleting RAD51. SRS2, given its role in inhibiting an early step in recombination [98–100], probably prevents dna2–2 sgs1Δ–derived lesions from entering the recombinational repair pathway. Thus, the interaction between Sgs1p and Dna2p is multipotential.
Rrm3p promotes DNA replication through non-nucleosomal protein–DNA complexes, including the rDNA RFB, Rap1p-binding sites in the telomere, inactive ARS sites, and centromere DNA [101–104]. Rrm3p may act on the same replication intermediates as Dna2p rather than on downstream toxic intermediates formed during repair of faulty replication, since rad51 mutations do not suppress the dna2–2 rrm3Δ synthetic lethality. Recently, rrm3Δ has been tested for synthetic lethality against a number of candidate gene deletions. Unlike dna2–2 rrm3Δ, most of the synthetically lethal combinations, such as rrm3Δ srs2Δ and rrm3Δ sgs1Δ, were suppressed by recombination mutants [47,75]. Thus, early replication intermediates cause cell lethality in dna2 rrm3Δ mutants, while recombination intermediates cause cell lethality in rrm3Δ sgs1Δ mutants. These intermediates may not involve FEN1, since rad27Δ rrm3Δ is not synthetically lethal [47]. Although both dna2 and rrm3Δ mutants show significant pausing in the rDNA, a fob1Δ mutation did not restore viability to the dna2–2 rrm3Δ mutant (Table 2), so there are additional sequences replicated by these genes. We have recently demonstrated that Dna2p can stimulate FEN1 cleavage of long flaps with secondary structure, but that the reaction is inefficient [12,105]. Since Rrm3p is a 5′ to 3′ helicase, Rrm3p is a candidate for a helicase that may aid Dna2p in flap processing.
DNA Repair
It does not appear that dna2 interacts with genes involved in nucleotide excision repair, consistent with the relative resistance of dna2 mutants to UV irradiation [16]. Base excision repair (long patch) involves all of the proteins also involved in OFP, and therefore a role for Dna2p in base excision repair would be supported by the interactions found in this work and the MMS sensitivity of dna2 mutants [8]. A role for Dna2p in an unidentified Rad52p-dependent pathway was discussed above.
The synthetic lethality of dna2 with each member of the Mre11p complex may contribute to emerging evidence that the Mre11p complex functions at the replication fork. First, the seven genes whose interactions overlapped most significantly with dna2 (see Figure 2) are also synthetically lethal with the genes of the Mre11p complex. Second, the Mre11p complex associates with chromatin primarily during S phase, and this association does not appear to require DSBs [106]. It has been suggested that the Mre11p complex assists sister chromatid association [106], and that this association is required for recombinational repair of DSBs during DNA replication. Another view derives from the fact that the Mre11p complex and Exo1p are both required for activation of the Rad53p checkpoint kinase after inhibition of replication by HU. This leads to the inference that the Mre11p complex and Exo1p may convert DSBs arising at stalled replication forks into single-stranded DNA, a signal for subsequent repair. Since replication forks stall in dna2 mutants, the synthetic lethality with mutations in Mre11p complex components or Exo1p could be explained by a failure to produce the single-stranded DNA signal. It is possible that DSBs that arise during normal DNA replication are repaired in Mre11p complex mutants, but are lethal in cells lacking both Mre11p and Dna2p [7]. Finally, or alternatively, the synthetic lethality of dna2 mre11Δ may indicate that DNA2 is involved in a telomere defect, as has been shown in S. pombe [48,107]. We cannot eliminate the possibility that the Dna2p/Mre11p interaction is involved in repair, but we note that dna2–2 and rad50–5, which is as deficient in repair as rad52Δ, are epistatic with respect to repair.
Sister Chromatid Cohesion
Replication forks in dna2 mutants pause at the RFB in the rDNA, where DSBs result [17,35,37]. Our current finding that deletion of FOB1 suppresses the synthetic lethality of dna2–2 ctf4Δ damage could be explained if Ctf4p-mediated sister chromatid cohesion is necessary to repair damage at the RFB due to defective Dna2p. This requirement for cohesion in DNA2 mutants is not limited to the rDNA, since the dna2–2 ctf4Δ fob1Δ strain is ts and radiation sensitive. Thus, the Dna2p deficiency must give rise to damage requiring cohesion for repair elsewhere in the chromosome as well. A role for cohesion in maintaining the replication fork during stalling or collapse is attractive since cohesion is required for efficient DSB repair [108]; a role for sister chromatid cohesion in preventing excessive sister chromatid exchange due to breaks at the RFB has been directly demonstrated [109]. We note that our analysis of replication in dna2–2 strains by two-dimensional gel electrophoresis indicates a high incidence of stalled replication forks at sequences throughout the rDNA, not limited to the RFB, suggesting general replication fork stalling in dna2 mutants and providing physical evidence of a more delocalized requirement for sister chromatid cohesion, perhaps throughout the chromosome [35,37].
Chromatin Remodeling, Disassembly, and Reassembly
During DNA replication, the chromatin in front of the replication fork is disassembled and then reassembled behind the fork. Our new findings add to recent findings from many sources that are providing the first insight into the molecular links between the replication machinery and chromatin dynamics. Dna2p interacts with both Asf1p and yFACT. The Asf1p/Dna2p interaction in chromatin assembly/remodeling is too ill-defined for further inferences at the moment. However, the allele-specific synthetic lethality between dna2 and pol1–1 suggests that Dna2p participates in the recently demonstrated interplay between Spt16p (a component of the FACT-like nucleosome reorganization factor), Ctf4p, and pol α [56]. The pol1–1 mutant protein fails to interact with Spt16p and shows altered temporal interaction with Ctf4p. The compromised association between Spt16p and pol α in the pol1–1 mutant is accompanied by a delay both in pol α recruitment to late origins and its release, leading to a slow S phase [56]. By adding a link between Dna2p and this particular aspect of pol α function, our results support the model of the Formosa lab that the yeast Spt16p complex is likely to be directly involved in DNA replication [52–54,110], as has also been suggested for the frog FACT complex [111]. The Spt16p remodeling complex may facilitate the movement of pol α and Dna2p through nucleosomes, as proposed for human FACT in transcription [112,113]. The caveat that yFACT defects might result in reduced transcription of gene(s) that interact with DNA2 was mentioned above.
Other DNA2-interacting genes encode specific sets of histone modification enzymes that catalyze histone ubiquitylation, methylation, and deacetylation. We discovered here that the synthetic sickness of dna2 and rad6 is related to Rad6p/Bre1p-mediated ubiquitylation of histone H2B, which in turn leads to methylation of H3 at the lysine at position 4 by the SET1 complex, containing Set1p, Swd1p, and Swd3p [114]. set1 mutants are sensitive to HU and may accumulate DNA damage during S phase. set1 was not recovered in the SGA screen because it is not in the deletion collection. Ubiquitylation and methylation alter silencing and chromatin structure at the rDNA and at telomeres, which may suggest a mechanism for interaction with dna2 [114,115].
The interaction between Dna2p and Hst3p and the Rpd3p/Pho23p histone deacetylases is interesting because mutations in ORC, the replication initiator, as well as rad27Δ, pol32Δ, and sgs1Δ also show synthetic interaction with hst3Δ [18,116]. HST3 performs a redundant function in DNA replication but hst3 hst4 cells have phenotypes indicative of replication defects, such as increased rates of chromosome loss and mitotic recombination, decreased telomere silencing, and hypersensitivity to UV treatment [117]. DNA2 is the only DNA replication gene thus far found to interact with RPD3. Histone deacetylation by the Rpd3p/Pho23p complex has been previously implicated in the temporal regulation of origin activation, but not elongation, in DNA replication [67,68]. The observation that hst1Δ, hst3Δ, rpd3Δ, bre1Δ, spt16Δ, cac2Δ, and vid21Δ are all linked to the network shown in Figure 2 suggests that the existing nucleosome structure has been optimized for high fidelity DNA replication (see Protocol S1).
Conclusions
We propose that the genome maintenance network is coordinated by physical interaction of the replication proteins with the complexes that carry out regulation and repair. A good deal of evidence supports this. Originally we found that Dna2p co-purified with FEN1 and later demonstrated the genetic interaction. We discovered a genetic interaction between Dna2p and BLM and WRN helicases—the human orthologs of Sgs1p—and then found that this genetic interaction also represented a direct physical interaction between Dna2p and those helicases [94]. Others found genetic interactions between DNA2 and RFA1, encoding the single-stranded DNA binding protein RPA, and similarly documented a physical interaction between the proteins Dna2p and RPA [13,118]. Indeed, most DNA replication proteins multitask and are components of many complexes. Pol δ and ɛ are involved in DNA repair complexes as well as replication complexes. Pol ɛ is, in addition, involved in the S phase checkpoint. FEN1 itself is involved in multiple reactions, including long patch base excision repair [15]. Though circumstantial, the resemblance between the basic topology of the genetic network described here and the proteomic network described in bacteria [76] further suggests the model that replication proteins physically coordinate repair and regulatory genome maintenance complexes. We anticipate that this model will be verified in detail when proteomic approaches in yeast yield the kind of comprehensive data that can already be obtained from genetic screens such as SGA.
Materials and Methods
Strains.
All strains used in the study are found in Table S4. The strains used in the SGA screen and subsequent verification were as follows: 4741 MATa his3Δ1 leu2Δ0 met15Δ0 ura30Δ; 4741 Mat α his3Δ1 leu2Δ0 met15Δ0 ura3Δ0; strain 4741dna2–1–6D (MB100) MATa dna2-1his3Δ1 leu2Δ0 met15Δ0 ura3Δ0; strain 4741dna2–2–11D (MB101) MATα dna2–2::LEU2 his3Δ1 leu2Δ0 met15Δ0 ura3Δ0; Y3656 can1Δ::MFA1pr-HIS3-Mfα1pr-LEU2 his3Δ1 leu2Δ0 met15Δ0 ura3Δ0; Y3656dna2–1; Y3656 dna2–1::URA3; Y3656dna2–2; Y3556 dna2–2::URA3; and W30W303 MATa ade2–1 can1–100 his3–11,15 leu2–3,112, trp1–1 ura3 RAD5. Y3656 is derived from strain 4741 as described [19]. Unless otherwise noted, all haploid deletion mutants used in this work were in strain 4741 and all double mutants were tested using 4741dna2–1 or Y3656dna2–2 as indicated.
SGA screen.
The SGA screen was performed as previously described [19]. Y3656dna2–1 and Y356dna2–2 were constructed for this study and used as query strains. SGA analysis was performed for each of the dna2 alleles. Genes that showed synthetic lethality or synthetic sickness in the primary screen were tested by standard tetrad dissection [119]. For the secondary tetrad analysis, new heterozygous diploids were constructed between MB100 (4741 dna2–1) and MB101 (4741 dna2–2) and each of the candidate deletion mutants in strain 4741 (Invitrogen, Carlsbad, California, United States). Thus, each synthetic interaction that is reported here was tested using two independent diploids, one in the original screen and one in the secondary screen. The dna2–2 dissections were incubated at 30 °C and dna2–1 dissections at 23 °C. Any strain failing to generate viable double mutants was deemed synthetically lethal. A strain was considered synthetically sick if tetrads gave fewer double mutants than expected and double mutants grew slower than either single mutant. At least ten tetrads were dissected for each double mutant, and usually more were dissected.
X-ray and MMS sensitivity.
The X-ray source was Pantak (East Haven, Connecticut, United States) MK II 70 kev 20 ma. The source was calibrated and experiments were carried out as previously described [16]. For the MMS sensitivity assay, 0.005% and 0.01% MMS was added to yeast-peptone-dextrose (YPD) plates after autoclaving, and plates were used the same day.
Supporting Information
Protocol S1 Other
(53 KB DOC)
Click here for additional data file.
Table S1 Genes That Show Little, If Any, Interaction with dna2–1 or dna2–2 in Secondary Screen
(107 KB PDF)
Click here for additional data file.
Table S2 Synthetic Lethal Interactions Used to Prepare Figure 2
(111 KB PDF)
Click here for additional data file.
Table S3 Number of Spores after Dissection of dna2–2 mec1 tel1 slm1 Heterozygote
(95 KB PDF)
Click here for additional data file.
Table S4 Strains Used
(103 KB PDF)
Click here for additional data file.
This work was supported in part by United States Public Health Service grant GM25508 to JLC and by a National Science and Engineering Research Council of Canada grant, a grant from the Canadian Institutes of Health Research, and funds from Genome Canada, the Ontario Genome Institute, and the Ontario Research and Development Challenge Fund to CB.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. MEB, AHYT, PP, CB, and JLC conceived and designed the experiments. MEB, AHYT, PP, and XP performed the experiments. MEB, AHYT, PP, CB, and JLC analyzed the data. MEB, CB, and JLC wrote the paper.
A previous version of this article appeared as an Early Online Release on October 12, 2005 (DOI: 10.1371/journal.pgen.0010061.eor).
Abbreviations
DSBdouble-strand break
FAA2-amino-5-fluorobenzoic acid
HUhydroxyurea
MMSmethyl methane sulfonate
OFPOkazaki fragment processing
polDNA polymerase
RFBreplication fork barrier
SGAsynthetic genetic array
tstemperature-sensitive
WTwild-type
YPDyeast-peptone-dextrose
==== Refs
References
Kuo CL Huang CH Campbell JL 1983 Isolation of yeast DNA replication mutants using permeabilized cells Proc Natl Acad Sci U S A 80 6465 6469 6356128
Budd ME Campbell JL 1995 A new yeast gene required for DNA replication encodes a protein with homology to DNA helicases Proc Natl Acad Sci U S A 92 7642 7646 7644470
Budd ME Choe WC Campbell JL 1995
DNA2 encodes a DNA helicase essential for replication of eukaryotic chromosomes J Biol Chem 270 26766 26769 7592912
Bae SH Choi E Lee K Park J Lee S 1998 Dna2 of Saccharomyces cerevisiae possesses a single-stranded DNA-specific endonuclease activity that is able to act on double-stranded DNA in the presence of ATP J Biol Chem 273 26880 26890 9756935
Waga S Bauer G Stillman B 1994 Reconstitution of complete SV40 DNA replication with purified replication factors J Biol Chem 269 10923 10934 8144677
Harrington JJ Lieber MR 1994 The characterization of a mammalian DNA structure-specific endonuclease EMBO J 13 1235 1246 8131753
Budd ME Campbell JL 1997 A yeast replicative helicase, Dna2 helicase, interacts with yeast FEN-1 nuclease in carrying out its essential function Mol Cell Biol 17 2136 2142 9121462
Formosa T Nittis T 1999 Dna2 mutants reveal interactions with DNA polymerase alpha and Ctf4, a Pol alpha accessory factor, and show that full DNA2 helicase activity is not essential for growth Genetics 151 1459 1470 10101169
Bae SH Seo YS 2000 Characterization of the enzymatic properties of the yeast Dna2 helicase/endonuclease suggests a new model for Okazaki fragment processing J Biol Chem 275 38022 38031 10984490
Ayyagari R Gomes XV Gordenin DA Burgers PMJ 2003 Okazaki fragment maturation in yeast. I. Distribution of functions between Fen1 and Dna2 J Biol Chem 278 1618 1625 12424238
Jin YH Ayyagari R Resnick MA Gordenin DA Burgers PMJ 2003 Okazaki fragment maturation in yeast. II. Cooperation between the polymerase and 3′ to 5′ exonuclease activities of Pol delta in the creation of a ligatable nick J Biol Chem 278 1626 1633 12424237
Kao HI Veeraraghavan J Polaczek P Campbell JL Bambara RA 2004 On the roles of Saccharomyces cerevisiae Dna2p and FEN1 in Okazaki fragment processing J Biol Chem 279 15014 15024 14747468
Bae SH Bae KH Kim JA Seo YS 2001 RPA governs endonuclease switching during processing of Okazaki fragments in eukaryotes Nature 412 456 461 11473323
Kao HI Campbell JL Bambara RA 2004 Dna2p helicase/nuclease is a tracking protein, like FEN1, for flap cleavage during Okazaki fragment maturation J Biol Chem 279 50840 50849 15448135
Toueille M Hubscher U 2004 Regulation of the DNA replication fork: A way to fight genomic instability Chromosoma 113 113 125 15300444
Budd ME Campbell JL 2000 The pattern of sensitivity of yeast dna2 mutants to DNA damaging agents suggests a role in DSB and postreplication repair pathways Mutat Res 459 173 186 10812329
Choe W Budd M Imamura O Hoopes L Campbell JL 2002 Dynamic localization of an Okazaki fragment processing protein suggests a novel role in telomere replication Mol Cell Biol 22 2002 2017
Tong AHY Lesage G Bader GD Ding H Xu H 2004 Global mapping of the yeast genetic interaction network Science 303 808 813 14764870
Tong AHY Evangelista M Parsons AB Xu Bader GD 2001 Systematic genetic analysis with ordered arrays of yeast deletion mutants Science 294 2364 2368 11743205
Hanna JS Kroll ES Lundblad V Spencer FA 2001
Saccharomyces cerevisiae CTF18 and CTF4 are required for sister chromatid cohesion Mol Cell Biol 21 3144 3158 11287619
Budd ME Choe WC Campbell JL 2000 The nuclease activity of the yeast Dna2 protein, which is related to the RecB-like nucleases, is essential in vivo J Biol Chem 275 16518 16529 10748138
Subramanya HS Bird LE Brannigan JA Wigley DB 1996 Crystal structure of a DExx box DNA helicase Nature 384 379 383 8934527
Araki Y Kawasaki Y Sasanuma H Tye BK Sugino A 2003 Budding yeast mcm10/dna43 mutant requires a novel repair pathway for viability Genes Cells 8 465 480 12694535
Ireland MJ Reinke SS Livingston DM 2000 The impact of lagging strand replication mutations on the stability of CAG repeat tracts in yeast Genetics 155 1657 1665 10924464
Frank G Braunshofer-Reiter C Wintersberger U 1998 Yeast RNase H(35) is the counterpart of the mammalian RNase HI and is evolutionarily related to prokaryotic RNase HII FEBS Lett 421 23 26 9462832
Johnson RE Kovvali GK Prakash L Prakash S 1998 Role of yeast Rth1 nuclease and its homologs in mutation avoidance, DNA repair, and DNA replication Curr Genet 34 21 29 9683672
Tishkoff DX Boerger AL Bertrand P Filosi N Gaida GM 1997 Identification and characterization of Saccharomyuces cerevisiae EXO1, a gene encoding an exonuclease that interacts with MSH2 Proc Natl Acad Sci U S A 94 7487 7492 9207118
Parenteau J Wellinger RJ 1999 Accumulation of single-stranded DNA and destabilization of telomeric repeats in yeast mutant strains carrying a deletion of RAD27
Mol Cell Biol 19 4143 4152 10330154
Tran PT Erdeniz N Dudley S Liskay RM 2002 Characterization of nuclease-dependent functions of Exo1p in Saccharomyces cerevisiae
DNA Repair (Amst) 1 895 912 12531018
Smolikov S Mazor Y Krauskopf A 2004 ELG1, a regulator of genome stability, has a role in telomere length regulation and in silencing Proc Natl Acad Sci U S A 101 1656 1661 14745004
Kanellis P Agyel R Durocher D 2003 Elg1 forms an alternative PCNA-interacting RFC complex required to maintain genome stability Curr Biol 13 1583 1595 13678589
Bellaoui M Chang M Ou J Xu H Boone C 2003 Elg1 forms an alternative RFC complex important for DNA replication and genome integrity EMBO J 22 4304 4313 12912927
Garg P Stith CM Sabouri N Johansson E Burgers PM 2004 Idling by DNA polymerase delta maintains a ligatable nick during lagging-strand DNA replication Genes Dev 18 2764 2773 15520275
Johansson E Garg P Burgers PMJ 2004 The Pol32 subunit of DNA polymerase delta contains separable domains for processive replication and proliferating cell nuclear antigen (PCNA) binding J Biol Chem 279 1907 1915 14594808
Weitao T Budd M Mays Hoopes LL Campbell JL 2003 Dna2 helicase/nuclease causes replicative fork stalling and double-strand breaks in the ribosomal DNA of Saccharomyces cerevisiae
J Biol Chem 278 22513 22522 12686542
Mays Hoopes LL Budd M Choe W Weitao T Campbell JL 2002 Mutations in DNA replication genes reduce yeast life span Mol Cell Biol 22 4136 4146 12024027
Weitao T Budd M Campbell JL 2003 Evidence that yeast SGS1, DNA2, SRS2, and FOB1 interact to maintain rDNA stability Mutat Res 532 157 172 14643435
Gangloff S Soustelle C Fabre F 2000 Homologous recombination is responsible for cell death in the absence of the Sgs1 and Srs2 helicases Nat Genet 25 192 194 10835635
Mullen JR Kaliraman V Ibrahim SS Brill SJ 2001 Requirement for three novel protein complexes in the absence of the Sgs1 DNA helicase in Saccharomyces cerevisiae
Genetics 157 103 118 11139495
Hryciw T Tang M Fontanie T Xiao W 2002
MMS1 protects against replication-dependent DNA damage in Saccharomyces cerevisiae
Mol Genet Genomics 266 848 857 11810260
Symington LS 2002 Role of RAD52 epistasis group genes in homologous recombination and double-strand break repair Microbiol Mol Biol Rev 66 630 670 12456786
Llorente B Symington LS 2004 The Mre11 nuclease is not required for 5′ to 3′ resection at multiple HO-induced double-strand breaks Mol Cell Biol 24 9682 9694 15485933
Fiorentino DF Crabtree GR 1997 Characterization of Saccharomyces cerevisiae dna2 mutants suggests a role for the helicase late in S phase Mol Biol Cell 8 2519 2537 9398673
Alcasabas AA Osborn AJ Bachant J Hu F Werler PJH 2001 Mrc1 transduces signals of DNA replication stress to activiate Rad53 Nat Cell Biol 3 958 965 11715016
Osborn AJ Elledge SJ 2003 Mrc1 is a replication fork component whose phosphorylation in response to DNA replication stress activates Rad53 Genes Dev 17 1755 1767 12865299
Katou Y Kanoh Y Bando M Noguchi H Tanaka H 2003 S-phase checkpoint proteins Tof1 and Mrc1 form a stable replication-pausing complex Nature 424 1078 1083 12944972
Torres JZ Schnakenberg SL Zakian VA 2004
Saccharomyces cerevisiae Rrm3p DNA helicase promotes genome integrity by preventing replication fork stalling: Viability of rrm3 cells requires the intra-S-phase checkpoint and fork restart activities Mol Cell Biol 24 3198 3212 15060144
Lisby M Barlow JH Burgess RC Rothstein R 2004 Choreography of the DNA damage response: Spatiotemporal relationships among checkpoint and repair proteins Cell 118 699 713 15369670
Ritchie KB Mallory JC Petes TD 1999 Interactions of TLC1 (which encodes the RNA subunit of telomerase), TEL1, and MEC1 in regulating telomere length in the yeast Saccharomyces cerevisiae
Mol Cell Biol 19 6065 6075 10454554
Miles J Formosa T 1992 Evidence that POB1, a Saccharomyces cerevisiae protein that binds to DNA polymerase alpha, acts in DNA metabolism in vivo Mol Cell Biol 12 5274 5735
Edwards SE Li CX Levy DL Brown J Snow PM 2003
Saccharomyces cervisiae DNA polymerase epsilon and polymerase sigma interact physically and functionally, suggesting a role for polymerase epsilon in sister chromatid cohesion Mol Cell Biol 23 2733 2748 12665575
Schlesinger MB Formosa T 2000
POB3 is required for both transcription and replication in the yeast Saccharomyces cerevisiae
Genetics 155 1593 1606 10924459
Formosa T Eriksson P Wittmeyer J Ginn J Yu Y 2001 Spt16-Pob3 and the HMG protein Nhp6 combine to form the nucleosome-binding factor SPN EMBO J 20 3506 3517 11432837
Rhoades AR Ruone S Formosa T 2004 Structural features of nucleosomes reorganized by yeast FACT and its HMG box component, Nhp6 Mol Cell Biol 24 3907 3917 15082784
Formosa T Ruone S Adams MD Olsen AE Eriksson P 2002 Defects in SPT16 or POB3 (yFACT) in Saccharomyces cerevisiae cause dependence on the Hir/Hpc pathway: Polymerase passage may degrade chromatin structure Genetics 162 1557 1571 12524332
Zhou Y Wang TSF 2004 A coordinated temporal interplay of nucleosome reorganization factor, sister chromatin cohesion factor, and DNA polymerase alpha facilitates DNA replication Mol Cell Biol 24 9568 9579 15485923
Budd M Campbell JL 1987 Temperature sensitive mutants of yeast DNA polymerase I Proc Natl Acad Sci U S A 84 2838 2842 3554248
Pellicioli A Lucca C Liberi G Marini F Lopes M 1999 Activation of Rad53 kinase in response to DNA damage and its effect in modulating phosphorylation of the lagging strand DNA polymerase EMBO J 18 6561 6572 10562568
Verreault A 2003 Histone deposition at the replication fork: A matter of urgency Mol Cell 11 283 284 12620214
Prakash L 1981 Characterization of postreplication repair in Saccharomyces cerevisia e and effects of rad6, rad18, rev3, and rad52 mutations Mol Gen Genet 184 471 478 7038396
Sung PE Berleth E Prakash S Prakash L 1991 Yeast RAD6 encoded ubiquitin conjugating enayme mediates protein degradation dependent on the N-end-recognizing E3 enzyme EMBO J 10 2187 2193 2065660
Wood A Krogan NJ Dover J Schneider J Heidt J 2003 Bre1, an E3 ubiquitin ligase required for recruitment and substrate selection of Rad6 at a promoter Mol Cell 11 267 274 12535539
Bailly V Lamb J Sung P Prakash S Prakash L 1994 Specific complex formation between yeast RAD6 and RAD18 proteins: A potential mechanism for targeting RAD6 ubiquitin-conjugating activity to DNA damage sites Genes Dev 8 811 820 7926769
Ulrich HD Jentsch S 2000 Two RING finger proteins mediate cooperation between ubiquitin-conjugating enzymes in DNA repair EMBO J 19 3388 3397 10880451
Hwang WW Venkatasubrahmanyam S Ianculescu AG Tong A Boone C 2003 A conserved RING finger protein required for histone H2B monoubiquitination and cell size control Mol Cell 11 261 266 12535538
Ng HH Dole S Struhl K 2003 The Rtf1 component of the Paf1 transcriptional elongation complex is required for ubiquitination of histone H2B J Biol Chem 278 33625 33628 12876293
Vogelauer M Rubbi L Lucas I Brewer BJ Grunstein M 2002 Histone acetylation regulates the time of replication origin firing Mol Cell 10 1223 1233 12453428
Aparicio JG Viggiani CJ Gibson DG Aparicio OM 2004 The Rpd3-Sin3 histone deacetylase regulates replication timing and enables intra-S origin control in Saccharomyces cerevisiae
Mol Cell Biol 24 4769 4780 15143171
Myung K Datta A Chen C Kolodner RD 2001
SGS1, the Saccharomyces cerevisiae homologue of BLM and WRN, suppresses genome instability and homeologous recombination Nat Genet 27 113 116 11138010
Myung K Datta A Kolodner RD 2001 Suppression of spontaneous chromosomal rearrangements by S phase checkpoint functions in Saccharomyces cerevisiae
Cell 104 397 408 11239397
Myung K Chen C Kolodner RD 2001 Multiple pathways cooperate in the suppression of genome instability in Saccharomyces cerevisiae
Nature 411 1073 1076 11429610
Myung K Pennaneach V Kats ES Kolodner RD 2003
Saccharomyces cerevisiae chromatin-assembly factors that act during DNA replication function in the maintenance of genome stability Proc Natl Acad Sci U S A 100 6640 6645 12750463
Myung K Smith S Kolodner RD 2004 Mitotic checkpoint function in the formation of gross chromosomal rearrangements in Saccharomyces cerevisiae
Proc Natl Acad Sci U S A 101 15980 15985 15514023
Pennaneach V Kolodner RD 2004 Recombination and the Tel1 and Mec1 checkpoints differentially effect genome rearrangements driven by telomere dysfunction in yeast Nat Genet 36 612 617 15133512
Schmidt KH Kolodner RD 2004 Requirement of Rrm3 helicase for repair of spontaneous DNA lesions in cells lacking Srs2 or Sgs1 helicase Mol Cell Biol 24 3213 3226 15060145
Butland G Peregrin-Alvarez JM Li JL Yang W Yang X 2005 Interaction network containing conserved and essential protein complexes in Escherichia coli
Nature 433 531 537 15690043
Qiu J Qian Y Frank P Wintersberger U Shen B 1999
Saccharomyces cerevisiae RNase H(35) functions in RNA primer removal during lagging-strand DNA synthesis, most efficiently in cooperation with Rad27 nuclease Mol Cell Biol 19 8361 8371 10567561
Huertas P Aguilera A 2003 Cotranscriptionally formed DNA:RNA hybrids mediate transcription elongation impairment and transcription-associated recombination Mol Cell 12 711 721 14527416
Askree SH Yehuda T Smolikov S Gurevich R Hawk J 2004 A genome-wide screen for Saccharomyces cerevisiae deletion mutants that affect telomere length Proc Natl Acad Sci U S A 101 8658 8663 15161972
Loog M Morgan DO 2005 Cyclin specificity in the phosphorylation of cyclin-dependent kinase substrates Nature 434 104 15744308
Jin YH Obert R Burgers PMJ Kunkel TA Resnick MA 2001 The 3′-5′ exonuclease of DNA polymerase delta can substitute for the 5′ flap endonuclease Rad27/Fen1 in processing Okazaki fragments and preventing genome instability Proc Natl Acad Sci U S A 98 5122 5127 11309502
Huang ME de Calignon A Nicolas A Galibert F 2000 POL32, a subunit of the Saccharomyces cerevisiae DNA polymerase delta, defines a link between DNA replication and the mutagenic bypass repair pathway Curr Genet 38 178 187 11126776
Tanaka H Ryu GH Seo YS MacNeill SA 2004 Genetics of lagging strand DNA synthesis and maturation in fission yeast: Suppression analysis links the Dna2-Cdc24 complex to DNA polymerase delta Nucleic Acids Res 32 6367 6377 15576681
Ricke RM Bielinsky AK 2004 Mcm10 regulates the stability and chromatin association of DNA polymerase-alpha Mol Cell 16 173 185 15494305
Fien K Cho YS Lee JK Raychaudhuri S Tappin I 2004 Primer utilization by DNA polymerase alpha-primase is influenced by its interaction with Mcm10p J Biol Chem 279 16144 16153 14766746
Loeillet S Palancade B Cartron M Thierry A Fichard GF 2005 Genetic network interactions among replication repair and nuclear pore deficiencies in yeast DNA Repair (Amst) 4 459 468 15725626
Tishkoff DX Filosi N Gaida GM Kolodner RD 1997 A novel mutation avoidance mechanism dependent on Saccharomyces cerevisiae RAD27 is distinct from DNA mismatch repair Cell 88 253 263 9008166
Liu Y Kao HI Bambara RA 2004 Flap endonuclease 1: A central component of DNA metabolism Annu Rev Biochem 73 589 615 15189154
Johnson RE Gopala KK Prakash L Prakash S 1995 Requirement for the yeast RTH1 5′ to 3′ exonuclease for the stability of simple repetitive DNA Science 269 238 240 7618086
Callahan JL Andrews KJ Zakian VA Freudenreich CH 2003 Mutations in yeast replication proteins that increase CAG/CTG expansions also increase repeat fragility Mol Cell Biol 23 7849 7860 14560028
Lopes J Debrauwere H Buard J Nicolas A 2002 Instability of the human minisatellite CEB1 in rad27 Δ and dna2–1 re plication-deficient yeast cells EMBO J 21 3201 3211 12065432
Ooi SL Shoemaker DD Boeke JD 2003 DNA helicase gene interaction network defined using synthetic lethality analyzed by microarray Nat Genet 35 277 286 14566339
Liberi G Maffioletti G Lucca C Chiolo I Baryshnikova A 2005 Rad51-dependent DNA structures accumulate at damaged replication forks in sgs1 mutants defective in the yeast ortholog of BLM RecQ helicase Genes Dev 19 339 350 15687257
Imamura O Campbell JL 2003 The human Bloom syndrome gene suppresses the DNA replication and repair defects of yeast dna2 mutants Proc Natl Acad Sci U S A 100 8193 8198 12826610
Sharma S Sommers JA Brosh RM Jr 2004 In vivo function of the conserved non-catalytic domain of Werner syndrome helicase in DNA replication Hum Mol Genet 13 2247 2261 15282207
Oakley TJ Hickson ID 2002 Defending genome integrity during S-phase: Putative roles for RecQ helicases and topoisomerases DNA Repair (Amst) 1 175 207 12509252
Courcelle J Hanawalt PC 1999 RecQ and RecJ process blocked replication forks prior to resumption of replication in UV-irradiated Escherichia coli
Mol Gen Genet 262 543 551 10589843
Schiestl RH Prakash S Prakash L 1990 The SRS2 suppressor of rad6 mutations of Saccharomyces cerevisiae acts by channeling DNA lesions into the RAD52 DNA repair pathway Genetics 124 817 831 2182387
Veaute X Jeusset J Soustelle C Kowalczykowski SC Le Cam E 2003 The Srs2 helicase prevents recombination by disruption Rad51 nucleorotein filaments Nature 423 309 312 12748645
Krejci L Van Komen S Li Y Villemain J Reddy MS 2003 DNA helicase Srs2 disrupts the Rad51 presynaptic filament Nature 423 305 309 12748644
Ivessa AS Zakian VA 2002 To fire or not to fire: Origin activation in Saccharomyces cerevisiae ribosomal DNA Genes Dev 16 2459 2464 12368256
Ivessa AS Zhou JQ Schulz VP Monson EK Zakian VA 2002
Saccharomyces Rrm3p, a 5′ to 3′ DNA helicase that promotes replication fork progression through telomeric and subtelomeric DNA Genes Dev 16 1383 1396 12050116
Ivessa AS Lenzmeier BA Bessler JB Goudsouzian LK Schnakenberg SL 2003 The Saccharomyces cerevisiae helicase Rrm3p facilitates replication past nonhistone protein-DNA complexes Mol Cell 12 1525 1536 14690605
Makovets S Herskowitz I Blackburn EH 2004 Anatomy and dynamics of DNA replication fork movement in yeast telomeric regions Mol Cell Biol 24 4019 4031 15082794
Kao HI Henricksen LA Liu Y Bambara RA 2002 Cleavage specificity of Saccharomyces cerevisiae flap endonuclease 1 suggests a double-flap structure as the cellular substrate J Biol Chem 277 14379 14389 11825897
Mirzoeva OK Petrini JHJ 2003 DNA replication-dependent nuclear dynamics of the Mre11 complex Mol Cancer Res 1 207 218 12556560
Tomita K Kibe T Kang HY Seo YS Uritani M 2004 Fission yeast Dna2 is required for generation of the telomeric single-strand overhang Mol Cell Biol 24 9557 9567 15485922
Sjogren C Nasmyth K 2001 Sister chromatid cohesion is required for postreplicative double-strand break repair in Saccharomyces cerevisiae
Curr Biol 11 991 995 11448778
Kobayashi T Horiuchi T Tongalnkar P Vu L Nomura M 2004 SIR2 regulates recombination between different rDNA repeats, but not recombination within individual rRNA genes in yeast Cell 117 441 453 15137938
Wittmeyer J Formosa T 1997 The Saccharomyces cerevisiae DNA polymerase alpha catalytic subunit interacts with Cdc68/Spt16 and with Pob3, a protein similar to an HMG1-like protein Mol Cell Biol 17 4178 4190 9199353
Okuhara K Ohta K Seo H Shioda M Yamada T 1999 A DNA unwinding factor involved in DNA replication in cell-free extracts of Xenopus eggs Curr Biol 9 341 350 10209116
Orphanides G Wu WH Lane WS Hampsey M Reinberg D 1999 The chromatin-specific transcrition elongation factor FACT comprises human ST16 and SSRP1 proteins Nature 400 284 288 10421373
Orphanides G LeRoy G Chang CH Luse DS Reinberg D 1998 FACT, a factor that facilitates transcript elongation through nucleosomes Cell 92 105 116 9489704
Nagy PL Griesenbeck J Kornberg RD Cleary ML 2002 A trithorax-group complex purified from Saccharomyces cerevisiae is required for methylation of histone H3 Proc Natl Acad Sci U S A 99 90 94 11752412
Krogan NJ Dover J Khorrami S Greenblatt JF Schneider J 2002 COMPASS, a histone H3 (lysine 4) methyltransferase required for telomeric silencing of gene expression J Biol Chem 277 10753 10755 11805083
Suter B Tong A Chang M Yu L Brown GW 2004 The origin recognition complex links replication, sister chromatid cohesion and transcriptional silencing in Saccharomyces cerevisiae
Genetics 167 579 591 15238513
Brachmann CB Sherman JM Devine SE Cameron EE Pillus L 1995 The SIR2 gene family, conserved from bacteria to humans, functions in silencing, cell cycle progression, and chromosome stability Genes Dev 9 2888 2902 7498786
Bae KH Kim HS Bae SH Kang HY Brill S 2003 Bimodal interaction between replication-protein A and Dna2 is critical for Dna2 function both in vivo and in vitro Nucleic Acids Res 31 3006 3015 12799426
Guthrie C Fink G 1991 Guide to yeast genetics and molecular biology New York Academic Press 933 p.
Morrison A Sugino A 1992 Roles of POL3, POL2 and PMS1 genes in maintaining accurate DNA replication Chromosoma 102 S147 S149 1291235
Toyn JH Gunyukzlu PL White WH Thompson LA Hollis GF 2000 A counterselection for the tryptohpan pathway in yeast: 5-fluoroanthranilic acid resistance Yeast 16 553 560 10790693
|
16327883
|
PMC1298934
|
CC BY
|
2021-01-05 07:59:37
|
no
|
PLoS Genet. 2005 Dec 2; 1(6):e61
|
utf-8
|
PLoS Genet
| 2,005 |
10.1371/journal.pgen.0010061
|
oa_comm
|
==== Front
PLoS GenetPLoS GenetpgenplgeplosgenPLoS Genetics1553-73901553-7404Public Library of Science San Francisco, USA 1632788410.1371/journal.pgen.001006405-PLGE-RA-0066R3plge-01-05-06Research ArticlePediatricsGenetics/Gene DiscoveryGenetics/Gene FunctionGenetics/Comparative GenomicsGenetics/Functional GenomicsGenetics/Genetics of DiseaseGenetics/Complex TraitsHomo (Human)Medical Sequencing of Candidate Genes for Nonsyndromic Cleft Lip and Palate Direct Sequencing of CL/P GenesVieira Alexandre R 1Avila Joseph R 1Daack-Hirsch Sandra 1Dragan Ecaterina 1Félix Têmis M 1Rahimov Fedik 2Harrington Jill 1Schultz Rebecca R 1Watanabe Yoriko 1Johnson Marla 1Fang Jennifer 1O'Brien Sarah E 1Orioli Iêda M 3Castilla Eduardo E 45FitzPatrick David R 6Jiang Rulang 7Marazita Mary L 89Murray Jeffrey C 1*1 Department of Pediatrics, University of Iowa, Iowa City, Iowa, United States of America
2 Interdisciplinary Genetics PhD Program, University of Iowa, Iowa City, Iowa, United States of America
3 Latin American Collaborative Study of Congenital Malformations, Department of Genetics, Federal University of Rio de Janeiro, Rio de Janeiro, Brazil
4 Latin American Collaborative Study of Congenital Malformations, Department of Genetics, Fiocruz, Rio de Janeiro, Brazil
5 Center for Medical Education and Investigation, Buenos Aires, Argentina
6 Cell and Molecular Genetics, Medical Research Council Human Genetics Unit, Edinburgh, United Kingdom
7 Center for Oral Biology and Department of Biomedical Genetics, University of Rochester, Rochester, New York, United States of America
8 Center for Craniofacial and Dental Genetics, School of Dental Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania, United States of America
9 Department of Human Genetics, Graduate School of Public Health, University of Pittsburgh, Pittsburgh, Pennsylvania, United States of America
van Heyningen Veronica EditorWestern General Hospital, United Kingdom* To whom correspondence should be addressed. E-mail: [email protected] 2005 2 12 2005 17 10 2005 1 6 e643 3 2005 17 10 2005 Copyright: © 2005 Vieira et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.Nonsyndromic or isolated cleft lip with or without cleft palate (CL/P) occurs in wide geographic distribution with an average birth prevalence of 1/700. We used direct sequencing as an approach to study candidate genes for CL/P. We report here the results of sequencing on 20 candidate genes for clefts in 184 cases with CL/P selected with an emphasis on severity and positive family history. Genes were selected based on expression patterns, animal models, and/or role in known human clefting syndromes. For seven genes with identified coding mutations that are potentially etiologic, we performed linkage disequilibrium studies as well in 501 family triads (affected child/mother/father). The recently reported MSX1 P147Q mutation was also studied in an additional 1,098 cleft cases. Selected missense mutations were screened in 1,064 controls from unrelated individuals on the Centre d'Étude du Polymorphisme Humain (CEPH) diversity cell line panel. Our aggregate data suggest that point mutations in these candidate genes are likely to contribute to 6% of isolated clefts, particularly those with more severe phenotypes (bilateral cleft of the lip with cleft palate). Additional cases, possibly due to microdeletions or isodisomy, were also detected and may contribute to clefts as well. Sequence analysis alone suggests that point mutations in FOXE1, GLI2, JAG2, LHX8, MSX1, MSX2, SATB2, SKI, SPRY2, and TBX10 may be rare causes of isolated cleft lip with or without cleft palate, and the linkage disequilibrium data support a larger, as yet unspecified, role for variants in or near MSX2, JAG2, and SKI. This study also illustrates the need to test large numbers of controls to distinguish rare polymorphic variants and prioritize functional studies for rare point mutations.
Synopsis
Nonsyndromic or isolated cleft lip with or without cleft palate (CL/P) is a birth defect with wide geographic distribution, occurring with an average frequency of 1/700 live births. Treatment can be provided, but it will involve medical, surgical, dental, and psychological personnel. Several different genes have been implicated in different cases. Here the researchers report the results of sequencing 20 different genes in 184 CL/P cases selected with an emphasis on more severe cases and cases with a positive family history for CL/P. Genes were selected based on previous work done by others and by the researchers' group. The authors' results suggest that point mutations in these candidate genes are likely to contribute to about 5% of CL/P, and particularly those with more severe phenotypes (bilateral cleft of the lip with cleft palate). This study also illustrates the need to test large numbers of controls to distinguish rare polymorphic variants and allow investigators to focus functional studies on the rare point mutations that seem to be disease-causing, so that researchers might better understand the mechanisms that play a role in CL/P.
Citation:Vieira AR, Avila JR, Daack-Hirsch S, Dragan E, Félix TM, et al. (2005) Medical sequencing of candidate genes for nonsyndromic cleft lip and palate. PLoS Genet 1(6): e64.
==== Body
Introduction
Nonsyndromic or isolated cleft lip with or without palate (CL/P) occurs in wide geographic distribution with an average birth prevalence of 1/700 [1]. CL/P is a complex trait determined by multiple, interacting loci, with additional environmental covariates. Recent work suggests that three to 14 interacting loci provide a good model for genetic effects in CL/P [2].
Studying candidate genes for CL/P selected from animal models and expression patterns is a common strategy [3]. To identify gene(s) involved in CL/P, investigators have used both association and linkage approaches to evaluate genetic contributions. To detect the very small gene effects on CL/P by linkage or linkage disequilibrium strategies, sample sizes need to be large and there needs to be either a common variant in association (for linkage disequilibrium) or a substantial degree of single locus contributions (for linkage). We used direct sequencing as an alternative approach to study candidate genes for CL/P hoping to identify genes with modest effects on the disease. The results of the direct sequencing of MSX1 [4,5] suggest that point mutations in this gene underlie approximately 2% of CL/P cases. We report here the results of sequencing 20 additional candidate genes for clefts. For seven genes with identified coding mutations that are potentially etiologic, we performed linkage disequilibrium studies as well. For the MSX1 P147Q mutation reported by Suzuki et al. [5], we investigated an additional 1,098 cleft cases.
Results
One hundred and forty-nine exons (representing 77,527 nucleotides of DNA sequencing), including exon–intron boundaries and untranslated regions, of 20 genes were screened for mutations in the Iowa and Philippines cleft populations. Table 2 summarizes the number of variants and putative mutations observed. Of the 256 variants seen, 16 missense mutations in nine genes seemed to be of potential etiologic importance. All 16 missense mutations were observed in a single cleft lip and palate case, with the exception of the SPRY2 D20A and TBX10 R354Q mutations that were seen in two and three cases respectively. None were seen in the 186 matched controls (Table 3). These mutation sites are not highly conserved across species with the exception of the SPRY2 and GLI2 mutations. Both SPRY2 mutation sites as well as three GLI2 mutation sites are conserved from Xenopus to human (Figure 1; complete data available at http://genetics.uiowa.edu/publication/html). The JAG2 and the TBX10 R354Q mutation sites are not conserved in other species orthologs available for study. The sequence surrounding the JAG2 A657H mutation site is likely a calcium-binding EGF-like domain, which is present in a large number of membrane-bound and extracellular proteins. Also, the SPRY2 K68N mutation site is in the sprouty domain and inhibits the Ras/mitogen-activated protein kinase (MAPK) cascade, a pathway crucial for developmental processes initiated by activation of various receptor tyrosine kinases.
Table 1 Candidate Genes Studied
Figure 1 Protein Comparisons of the Available Gene Orthologs for GLI2 S1213Y and SPRY2 D20A
GLI2 S1213Y (A) and SPRY2 D20A (B): Green bars indicate degree of conservation in each site. Amino acids in red indicate the mutation sites. All mutation comparisons are available as supplemental material at http://genetics.uiowa.edu/publications.html.
All mutations were predicted to be benign by PolyPhen (http://www.bork.embl-heidelberg.de/PolyPhen/) with the exception of the JAG2 M597I and SPRY2 D20A that were “possibly damaging” and GLI2 S1213Y that was “probably damaging” (Table 3). However, with the exception of the LHX8 E221A, GLI2 R426Q, and GLI2 S1213Y mutations, all missense mutations appear to potentially disrupt splicing by either creating or inactivating exonic splicing enhancer sequences (complete information is available as supplemental material at http://genetics.uiowa.edu/publications.html/). None of the mutations identified in this study appear to disrupt possible exonic splicing silencer sequences.
The SATB2 T190A mutation was not found in the panel of 1064 CEPH controls as well. We also tested the LHX8 E221A, SKI A388V, SPRY2 D20A, and TBX10 R354Q mutations in the panel of 1064 CEPH controls after not seeing it in 200 population matched controls. We found the LHX8 E221A mutation in 17 samples, the SKI A388V mutation in nine samples, the SPRY2 D20A mutation in 60 samples, and the TBX10 R354Q mutation in six samples. (A complete list is available at our Web site: http://genetics.uiowa.edu/publications.html).
The MSX1 P147Q mutation was not found in any of 1,671 controls but was found in two Filipino cleft families from a panel of 1,468 cleft cases from the Philippines, which indicates a frequency of 0.14%. The first family has no family history for clefting and the variant segregates from the unaffected mother. The second family has four affected with clefts. The variant was found in two cousins and segregates from the unaffected grandmother to her unaffected son and daughter of a sibship of nine. A first cousin of the proband is affected but does not carry the variant. The last affected, a third cousin once removed, also has a cleft but does not carry the variant (Figure 2).
Figure 2 Family Pedigree of the Filipino Family Segregating the MSX1 P147Q Mutation
PP (proline–proline) indicates the wild-type genotype. PQ (proline–glutamine) indicates the genotype of the individuals carrying the mutation.
The LHX8 and SATB2 mutations were originally seen in single cases from the Philippines. For SATB2 T190A, the variant was transmitted from the unaffected mother. The LHX8 E221A mutation was also transmitted from the unaffected mother and is present in one affected and one unaffected person but was absent in two unaffected siblings from a sibship of eight brothers and sisters (Figure 3). Both mutation sites are conserved between human and mouse (see Appendix at http://genetics.uiowa.edu/publications.html/). The SKI A388V mutation was seen in a case from Iowa. The mutation was transmitted from the mother and the mutation site is conserved between humans and species of fish and frog. Two cases from Iowa presented with the SPRY2 D20A mutation. Of these, one had parental DNA samples available and the mutation segregated from the unaffected mother. Three cleft individuals from Iowa presented the TBX10 R354Q mutation. In the two cases with parent samples available, one received the mutated allele from the mother and the other from the father (both unaffected).
Figure 3 Family Pedigree of the Filipino Proband with the LHX8 E221A Mutation
The segregation pattern of the mutation is not consistent with a simple Mendelian model for the disease.
EE (glutamic acid–glutamic acid) indicates the wild-type genotype. EA (glutamic acid–alanine) indicates the genotype of the individuals carrying the mutation.
Two additional interesting observations were made. An Iowa case presented with isolated cleft lip with cleft palate, and no family history of clefts or any features of DiGeorge syndrome was found to be homozygous for an intronic variant in the TBX1 gene (189 nucleotides into intron 8). This variant was not present in 186 matched controls. We tested this case for the presence of two copies of the ubiquitin fusion degradation gene (UFD1L), using an assay for DiGeorge or 22q syndrome [6], and the results were normal. Parental samples are not available to further study this case, nor is there enough DNA available to confirm a possible deletion by Southern blot analysis. The finding of a single rare homozygote variant suggests the possibility of a microdeletion or segmental isodisomy of this region. We tested four microsatellite markers (D22S420, D22S1685, D22S683, and D22S445) on both proband and mother samples. The mother presents distinct genotypes from the proband for D22S420 and D22S683 markers, however this finding does not exclude a segmental maternal isodisomy because the interval between these two markers, which contains TBX1, is 32 cM (data not shown).
The second observation involved four Filipino cases that presented a missense mutation in the last SKI amino acid (P728L). This mutation was not found in 186 matched controls. One case appears to have a de novo SKI P728L mutation. This case presented with an isolated right cleft lip and cleft palate and positive family history for clefts. Neither parent had this variant and their biological relationship to the case was confirmed after testing twenty polymorphic markers. Of the other three cases, one had a positive family history for clefts. For this case, we tested the parents, the paternal grandparents and three siblings out of a 15 sibship progeny. One of the tested siblings is affected with a right cleft lip and cleft palate associated with microcephaly. The SKI P728L variant segregated from the unaffected grandfather to the unaffected father. Of the three siblings tested, the two unaffected siblings had the variant, but the affected sibling did not. Based on this family, we concluded that this SKI P728L variant is probably not an etiologic mutation and included it in the column of non-synonymous coding variants in Table 2.
Table 2 Summary of Variants Found by Direct Sequence
Table 3 Potential Mutations Found in the Present Study
Linkage disequilibrium studies were performed for the genes GLI2, JAG2, MSX2, SATB2, SKI, SPRY2, and TBX10 in which likely etiologic missense mutations had been observed (complete information is available as supplemental material at http://genetics.uiowa.edu/publications.html/; results for FOXE1 were previously reported in Marazita et al. [7]). No single nucleotide polymorphism tested showed evidence for deviation from Hardy-Weinberg equilibrium in either the affected or unaffected individuals (data not shown). The haplotype analysis using the HBAT function of FBAT (http://www.biostat.harvard.edu/~fbat/fbat.htm) demonstrated borderline associations between MSX2 in both Filipino (p = 0.001) and Iowa (p = 0.008) populations, between JAG2 and the Filipinos (p = 0.004), and between both SATB2 (p = 0.03) and TBX10 (p = 0.04) and the Iowa population. However, when we combined the MSX2 haplotype data for both Filipino and Iowa populations, the association was weaker (p = 0.09). We also observed an association between CL/P and snp2 (rs2843159) in SKI (p = 0.000004) in the Filipino population. This association between CL/P and SKI in Filipinos was also suggested by the haplotype analysis (p = 0.0002). In addition, the same SKI snp2 marker showed association with cleft lip only in the South American clefting population from the Latin American Collaborative Study of Congenital Malformations (ECLAMC) (p = 0.004).
Discussion
Point mutations in the candidate genes FOXE1, GLI2, MSX2, SKI, SATB2, and SPRY2 appear in aggregate to contribute to as much as 6% of isolated cleft lip and palate cases, enriched for cases with bilateral cleft of the lip with cleft palate and a positive family history. The mutations found in this study are conserved in other mammals, may disrupt exonic splicing enhancer sequences, and were not found in between 400 to 2,000 control chromosomes. The JAG2 M597I and A657H mutations, although they appear to disrupt exonic splicing enhancer sequences and possibly damage the JAG2 protein, according to the PolyPhen prediction, are not conserved in other species and may be rare polymorphic sites.
Testing a larger number of control samples proved to be a useful way to differentiate rare polymorphisms from etiologic mutations. The LHX8 E221A, SATB2 T190A, SKI A388V, SPRY2 D20A, and TBX10 R354Q variants initially were not observed in approximately 200 matched controls. This number of controls is commonly used to assume that if a variant is not present, it is likely causal despite models that have shown that a larger number of controls is useful in eliminating rare variants [8]. However, when we tested these variants in the extended set of 1,064 controls, we found the LHX8 E221A variant in 17 individuals, the SKI A388V mutation in nine, the SPRY2 D20A mutation in 60 individuals, and the TBX10 R354Q in six individuals. Although the presence of the amino acid changes in unaffected controls does not exclude them from playing a role in CL/P, it does place them in a lower priority group for additional functional analysis and makes them difficult to use in any applied genetic counseling setting as they may be, at best, modifiers with low penetrance contributory alleles.
Some variants, such as the SKI A388V mutation, which is conserved back to Xenopus and Tilapia but is found in controls, demonstrate that species conservation alone maybe not enough to argue for an etiologic role of a given variant.
Our study illustrates the difficulties of defining as causal a mutation rarely seen in the population. Many of the missense mutations found in the cases studied were seen only when we extended our screen to a larger control group comprised of samples from ethnically diverse groups from almost all parts of the world. In addition, none of the mutations segregated from an affected parent. Incomplete penetrance is likely the explanation for the mutations that may be causal as has been clearly shown for other genes that contribute to clefting such as MSX1 [9] and FGFR1 [10] mutations. It is likely that we found these mutations only because we tested cases more likely to present a stronger genetic contribution (cases with positive family history and bilateral cleft of the lip with cleft palate). Mutations like the MSX1 P147Q and others that appear to show incomplete penetrance are comparable to autosomal dominant disorders resulting from mutations in SCN5A [11], IRF6 [12], or NKX2.5 [13].
The MSX1 P147Q mutation was seen in two cleft cases from the Philippines, but in none of the over 1,600 controls. It appears that this specific mutation underlies approximately 0.15% cases of apparent isolated CL/P. As shown previously this variant results in variable expression and decreased penetrance that make prospective studies of its phenotypic outcome necessary before accurate genetic counseling risk can be measured [5].
Rigorous demonstration that a mutation disrupts a genuine exonic splicing enhancer requires that the sequence autonomously promote splicing and that enhancement be absent in the mutant. An advantage of the score matrix approach [14] is that it allows direct testing of predicted effects on individual putative enhancer sites, rather than having to characterize exonic splicing enhancers by testing multiple random mutations and/or deletions along an exon. All 15 mutations on Table S2 appear to inactivate and/or create a predicted exonic splicing enhancer of at least one of the four serine/arginine-rich (SR) proteins. However, the presence of a high-score motif in a sequence does not necessarily identify that sequence as an exonic splicing enhancer in its native context.
In TBX1, we found one rare intronic variant in homozygous form in an Iowa cleft case, which could indicate that clefts arise from recessive functional intronic mutations in TBX1, or microdeletions that cannot be visualized by direct sequencing. This case does not have a 22q deletion involving the UFD1L gene. Detecting this rare homozygote in the absence of this variant in any other of the 400 people tested suggests these individuals may be identical by descent at this locus and gene. This variant itself, or others in linkage disequilibrium in TBX1, might be a hypomorphic allele whose joint presence results in enough change in gene expression or function to trigger a phenotype. Therefore, other alleles in regulatory regions of TBX1 should be a priority for identification.
Previous work from our group has screened FGFR1, IRF6, MSX1, TGFA, and TGFB3 for mutations on cleft cases. Point mutations in MSX1 appear to contribute approximately to 2% of all CL/P cases [4]. FGFR1 point mutations also appear to contribute to CL/P. In addition, FGFR1 loss-of-function mutations can cause forms of Kallmann syndrome that mimic isolated CL/P at birth and during childhood. Mutations in IRF6, which cause the syndromic forms of clefts, the Van der Woude and popliteal pterygium syndromes [12], were not found in the same collection of cleft cases as the present study, although rare non-coding variants in conserved regions were disclosed. However, IRF6 is strongly associated with CL/P, and it is likely a genetic modifier for clefts [15,16]. Previously for TGFA, five variants in conserved non-coding segments were found in individual cases but not seen in 278 controls [17]. In the present study we found another nine non-coding rare variants in single individuals, but we did not find the original five reported rare variants or any coding mutation that could be etiologic. If these variants are disease-causing mutations, they could explain, not only the conflicting results from association studies of isolated orofacial clefts and TGFA variants [18], but also the linkage studies that suggest a cleft susceptibility loci in 2p13, the TGFA locus [7]. For TGFB3, we previously reported one missense mutation (K130R) in a cleft palate–only case not seen in 350 controls [19]. We did not find this or any other mutations in TGFB3 in our current study population.
Loss-of-function mutations in GLI2 are associated with pituitary anomalies and holoprosencephaly-like features [20]. In this report, the three pedigrees segregating GLI2 loss-of-function mutations with complete clinical information presented orofacial clefts and polydactyly. We found four missense mutations in GLI2 in highly conserved amino acids. One of the cases also presented with polydactyly (Table 3).
We performed linkage disequilibrium studies in the genes that we found potentially disease-causing missense mutations. We found association between a marker (rs2843159) and a haplotype in SKI in the Filipino population. This association was also found in an independent population dataset from South America. The SKI locus, 1p36.3, was previously suggested as a cleft susceptibility loci in Caucasians [21,22]. Ski null mice present with clefts involving the lip [22], and the association we found appears to be stronger when cases with the involvement of the lip are included.
The possible trend for an association between clefts with a palate phenotype and SATB2 are in agreement with the cytogenetic evidence. Deletions and balanced translocations point to the existence of a locus on 2q32-q33, for which haploinsufficiency results in isolated cleft palate. A mutation analysis of SATB2 (located at 2q32-q33) in 70 unrelated patients with isolated cleft palate only did not reveal any coding region variants [23]. However, a meta-analysis of 13 genome scans for clefts indicated 2q32-q35 as a clefting susceptibility locus [7]. We studied 184 cleft lip and palate cases and found one missense mutation in SATB2 (T190A) that was not seen in approximately 1,200 controls. Based on the linkage disequilibrium and mutation analysis results of our study, we believe a regulatory element outside SATB2 coding regions may be implicated in clefting.
In summary, point mutations in six of the 20 candidate genes selected from expression, animal, and human data may to contribute to about 5% of isolated clefts, more likely those with more-severe phenotypes and/or a positive family history. Etiologic variants in regulatory elements of SKI, JAG2, and MSX2 may contribute to isolated clefts as well. Predictions by ESEfinder (http://rulai.cshl.edu/tools/ESE/) and PolyPhen regarding the function of the missense mutations found in this study, as well as exonic splicing enhancement and protein damaging, are challenging to interpret.
A major challenge in these studies was the frequent absence of a cleft phenotype in near relatives of an affected proband with a cleft and a rare missense mutation. In some cases these variants may not be etiologic, but in others, reduced penetrance for the cleft may be an active force as has been seen commonly in clefts [9,10] and other birth defects such as congenital hearth disease. Similarly these mutations may only be modifiers of the phenotype.
Cases due to microdeletions or isodisomy may contribute to clefts as well. This study illustrates the validity of testing greater numbers of controls to determine rare polymorphic variants and prioritize functional studies for rare point mutations. Given other recent data on the roles of FGFR1, IRF6, and MSX1 in isolated CL/P, one can begin to consider sequencing of a panel of high-probability candidate genes for genetic counseling indication. Although issues of penetrance and even etiology for any given mutation are not yet resolved, progress in this direction is now measurable.
Materials and Methods
Two collections of CL/P cases, 91 from the Philippines and 93 from Iowa, United States, were used to search for mutations in 20 candidate genes (Table 1). We selected the more-severe cases from those available to us, and the sequenced samples were enriched by bilateral cleft lip and palate cases with a positive family history for clefts (39/91 from the Philippines and 16/93 from Iowa). Two cases were later found to have associated features—one with Stickler syndrome in which family history was initially not available, and a second with polydactyly. Cases from the Philippines were studied under the auspices of Operation Smile International [24]. Patients were seen and examined by a board certified clinical geneticist (JCM or colleagues; see Acknowledgments) at one of four sites within the Philippines (Cavite, Kalibo, Cebu, and Negros). Iowa cases were collected through the Iowa Birth Defects Registry [25]. Signed consents were obtained from all participants before a blood sample was obtained. DNA was extracted according to published protocols.
For each of the 20 candidate genes, all exons and 5′ and 3′ untranslated regions were sequenced in both directions. Primer sequences and PCR conditions are available on our Web site (http://genetics.uiowa.edu/publications.html). Primers for FOXE1, GLI2, MSX2, OSR2, and TGFBR1 were obtained from the literature [20,26–29]. Cycle sequencing was performed in a 20-μl reaction using 4 μl of Applied Biosystems Big Dye Terminator sequencing reagent, 1 μl of 5 μM sequencing primer, 1 μl of DMSO, 4 μl of 2.5× Buffer, and 2.5 ng/100 base pair of DNA template. Following a denaturation step at 96 °C for 30 s, reactions were cycle sequenced at 96 °C for 10 s, 50 °C for 5 s, and 60 °C for 4 min for 40 cycles. Cleanup was performed using standard protocols. Samples were resuspended in 40 to 100 μl of ddH2O, and 2.5 μl were then injected on an Applied Biosystems 3700 sequencer. The Applied Biosystems sequence software (version 2.1.2) was used for lane tracking and first pass base calling. Chromatograms were transferred to a Unix workstation, base called with PHRED (version 0.961028), assembled with PHRAP (version 0.960731), scanned by POLYPHRED (version 0.970312), and the results viewed with CONSED (version 4.0) [30]. When the results indicated a possible new variant, the case sample was resequenced, as well as other available family members, and population controls. These data were analyzed using the same method.
For any coding variant, we performed direct sequencing in 186 population-matched controls. Control populations were collected as described above for the cases and consisted of individuals with no CL/P or other recognized birth defect from adults at the same sites where cases were collected. If the variant was found in one or more controls, it was considered a polymorphism. To expand the number of controls tested, we developed allele-specific assays for the LHX8 E221A, SATB2 T190A, SKI A388V, SPRY2 D20A, and TBX10 R354Q mutations. We tested them in the CEPH Diversity Cell Line Panel [31], which is comprised of 1,064 DNA samples from cultured lymphoblastoid cell lines derived from individuals representing 51 different human populations.
We also developed an assay for the MSX1 P147Q missense mutation, recently reported in three Vietnamese cleft families [5]. Besides the 1,064 CEPH Diversity Cell Line Panel controls, we tested the MSX1 P147Q assay in an additional 607 Filipino controls and in a collection of 1,468 cleft cases from the Philippines as well.
For the nine genes with potentially etiologic missense mutations, we identified orthologs through BLAST search of the non-redundant database using Homo sapiens FOXE1, GLI2, LHX8, JAG2, MSX2, SATB2, SKI, SPRY2, and TBX10, as reference sequences. We performed protein sequence comparisons with the available species. We also used the ESEfinder software available online to predict the presence of exonic splicing enhancers [14], which appear to be prevalent, and may be present in most, if not all exons [32,33]. We screened the 141 exonic splicing silencer decamers that were identified by Wang et al. [34] to check if any of those could be affected by the missense mutations we found. Finally, we used the PolyPhen software, also available online, to predict the impact of the amino acid substitutions identified on the structure and function of the human protein [35–37].
Two single nucleotide polymorphisms in weak linkage disequilibrium with each other were selected for each population to perform linkage disequilibrium studies in the genes with missense mutations in cases but not in controls. Four single nucleotide polymorphisms were chosen for GLI2 based on the International HapMap Project's linkage disequilibrium pattern of the gene (data not shown). Frequency of the alleles can be found in the supplemental material (http://genetics.uiowa.edu/publications.html). TaqMan-based assays [38] were performed on Applied Biosystems 7900 HT Sequence Detection System (Applied Biosystems, Foster City, California, United States). For one marker in SKI (rs2843159), we used a kinetic polymerase chain reaction assay previously reported [39]. These linkage disequilibrium studies were composed of 296 complete triads (mother/father/affected child) from the Philippines and 205 from Iowa. These samples were obtained as described above for cases and controls investigated by sequencing. The Family Based Association Test (FBAT) [40–42] program was used for this analysis. Significance figures were accounted for using Bonferroni correction taking into account the number of tests carried out [43]. With the Bonferroni correction, alpha is 0.0003 (0.05/192 comparisons) for the individual marker analysis and 0.0001 (0.05/384 comparisons) for the haplotype analysis. Linkage disequilibrium studies for FOXE1 were previously reported in Marazita et al. [7].
A third clefting population sample set of 434 case/mother pairs from South America was used to replicate any significant association. These population samples are derived from ECLAMC, which is a hospital-based birth defects registry study that includes sites in Argentina, Bolivia, Brazil, Chile, Ecuador, Paraguay, Uruguay, and Venezuela. This study population has previously been described in detail [44,45]. To analyze the ECLAMC samples, the likelihood ratio test (LRT) of Weinberg [46] was applied under the assumption that the distribution of paternal alleles is the same as maternal.
Supporting Information
Figure S1 Protein Comparisons of the Available Gene Orthologs for the Mutations Found in the Present Study
Green bars indicate degree of conservation in each site. Amino acids in red indicate the mutation sites. All mutation comparisons are available as supplemental material at http://genetics.uiowa.edu/publications.html.
(56 KB PDF)
Click here for additional data file.
Table S1 Primers and PCR Conditions
(51 KB PDF)
Click here for additional data file.
Table S2 Exonic Splicing Enhancer Prediction Analysis for the Missense Mutations Found in the Present Study. Nucleotides in red indicate the mutation sites.
(61 KB PDF)
Click here for additional data file.
Table S3 Mutations Screened in the CEPH Diversity Cell Line Panel
(48 KB PDF)
Click here for additional data file.
Table S4 Markers Selected for Linkage Disequilibrium Studies
(49 KB PDF)
Click here for additional data file.
Table S5 Haplotype Frequencies
(46 KB PDF)
Click here for additional data file.
Table S6 Linkage Disequilibrium Studies of Candidate Genes for Clefting
(50 KB PDF)
Click here for additional data file.
Table S7 Haplotype Analysis
(54 KB PDF)
Click here for additional data file.
Table S8 SKI LRT Results for the South American (ECLAMC) Clefting Samples
(40 KB PDF)
Click here for additional data file.
Accession Numbers
The National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/) Unigene accession numbers for the genes discussed in this paper are Homo sapiens FOXE1 (NM_004473), GLI2 (NP_084655), JAG2 (NM_002226), LHX8 (AY449521), MSX2 (NM_002449), SATB2 (NM_015265), SKI (AY334556), SPRY2 (NM_ 005842), and TBX10 (AY229977).
We are indebted to all the families that participated in this project. Buena Nepomucena and Paul Romitti supported the sample collection. Diana Caprau provided sample management. Carla Nishimura, Sally Santiago, John Allaman, Bonnie Ludwig, Kristin Aquilino, Kristin Orr, Amy Mach, and Darren Schipper provided technical support. Clinical exams in the Philippines were carried out on some cases by Howard Saal and Chinto Fong. Nasir Malik and Heiner Westphal assisted with the LHX8 studies. Maurice Payne completed preliminary work on ISGF3G. Marianne Timm and Renata Lucia Ferreira de Lima did preliminary work on JAG2. Mark Kresowik did preliminary work on MSX2 and TBX1. Elliott Hill helped with TGFB3 sequencing. Satoshi Suzuki assisted with the MSX1 P147Q assay. Margaret Cooper helped with the LRT analysis. Financial support was provided by National Institutes of Health grants DE08559, DE16215, ES10876, and 5 D43 TW05503.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. ARV, IMO, EEC, and JCM conceived and designed the experiments. ARV, JRA, ED, TMF, FR, JH, RRS, YW, MJ, JF, and SEO performed the experiments. ARV, JRA, SDH, ED, TMF, FR, JH, RRS, YW, MJ, JF, SEO, IMO, EEC, DRF, RJ, MLM, and JCM analyzed the data. SDH, DRF, RJ, MLM, and JCM contributed reagents/materials/analysis tools. ARV and JCM wrote the paper.
A previous version of this article appeared as an Early Online Release on October 17, 2005 (DOI: 10.1371/journal.pgen.0010064.eor).
Abbreviations
CEPHCentre d'Étude du Polymorphisme Humain
CL/Pcleft lip with or without cleft palate
ECLAMCLatin American Collaborative Study of Congenital Malformations
LRTlikelihood ratio test
==== Refs
References
Mossey PA Little J 2002 Epidemiology of oral clefts: An international perspective Wyszynski DF Cleft lip and palate. From origin to treatment New York Oxford Press 127 158
Schliekelman P Slatkin M 2002 Multiplex relative risk and estimation of the number of loci underlying an inherited disease Am J Hum Genet 71 1369 1385 12454800
Murray JC Schutte BC 2004 Cleft palate: Players, pathways, and pursuits J Clin Invest 113 1676 1678 15199400
Jezewski PA Vieira AR Nishimura C Ludwig B Johnson M 2003 Complete sequencing shows a role for MSX1 in non-syndromic cleft lip and palate J Med Genet 40 399 407 12807959
Suzuki Y Jezewski PA Machida J Watanabe Y Shi M 2004 In a Vietnamese population, MSX1 variants contribute to cleft lip and palate Genet Med 6 117 125 15354328
Kariyazono H Ohno T Ihara K Igarashi H Joh-o K 2001 Rapid detection of the 22q11.2 deletion with quantitative real-time PCR Mol Cellular Probes 15 71 73
Marazita ML Murray JC Lidral AC Arcos-Burgos M Cooper ME 2004 Meta-analysis of 13 genome scans reveals multiple cleft lip/palate genes with novel loci on 9q21 and 2p32–35 Am J Hum Genet 75 161 173 15185170
Collins JS Schwartz CE 2002 Detecting polymorphisms and mutations in candidate genes Am J Hum Genet 71 1251 1252 12452182
van den Boogard MJH Dorland M Beemer FA van Amstel HKP 2000 MSX1 mutation is associated with orofacial clefting and tooth agenesis in humans Nat Genet 24 342 343 10742093
Dodé C Levilliers J Dupont JM De Paepe A Le Dû N 2003 Loss-of-function mutations in FGFR1 cause autosomal dominant Kallmann syndrome Nat Genet 33 463 465 12627230
Hong K Brugada J Oliva A Berruezo-Sanchez A Potenza D 2004 Value of electrocardiographic parameters and ajmaline test in the diagnosis of Brugada syndrome caused by SCN5A mutations Circulation 110 3023 3027 15520322
Kondo S Schutte BC Richardson RJ Bjork BC Knight AS 2002 Mutations in interferon regulatory factor 6 cause Van der Woude and popliteal pterygium syndromes Nat Genet 32 285 289 12219090
Watanabe Y Benson DW Yano S Akagi T Yoshino M 2002 Two novel frameshift mutations in NKX2.5 result in novel features including visceral inversus and sinus venosus type ASD J Med Genet 39 807 811 12414819
Cartegni L Wang J Zhu Z Zhang MQ Krainer AR 2003 ESEfinder: A web resource to identify exonic splicing enhancers Nucleic Acid Res 31 3568 3571 12824367
Zucchero T Cooper ME Maher BS Daack-Hirsch S Nepomuceno B 2004 Interferon regulatory factor (IRF6) is a modifier for isolated cleft lip and palate New Engl J Med 351 769 780 15317890
Scapoli L Palmieri A Martinelli M Pezetti F Carinci P 2005 Strong evidence of linkage disequilibrium between polymorphisms at the IRF6 locus and nonsyndromic cleft lip with or without cleft palate, in an Italian population Am J Hum Genet 76 180 183 15558496
Machida J Yoshiura K Funkhauser CD Natsume N Kawai T 1999 Transforming growth factor-α (TGFA): Genomic structure, boundary sequences, and mutation analysis in nonsyndromic cleft lip/palate and cleft palate only Genomics 61 237 242 10552925
Mitchell LE 1997 Transforming growth factor α locus and nonsyndromic cleft lip with or without cleft palate: A reappraisal Genet Epidemiol 14 231 240 9181353
Lidral AC Romitti PA Basart AM Doetschman T Leysens NJ 1998 Association of MSX1 and TGFB3 with nonsyndromic clefting in humans Am J Hum Genet 63 557 568 9683588
Roessler E Du YZ Mullor JL Casas E Allen WP 2003 Loss-of-function mutations in the human GLI2 gene are associated with pituitary anomalies and holoprosencephaly-like features Proc Natl Acad Sci U S A 100 13424 13429 14581620
Prescott NJ Lees MM Winter RM Malcolm S 2000 Identification of susceptibility loci for nonsyndromic cleft lip with or without cleft palate in a two stage genome scan of affected sib-pairs Hum Genet 106 345 350 10798365
Colmenares C Heilstedt HA Shaffer LG Schwartz S Berk M 2002 Loss of the SKI proto-oncogene in individuals affected with 1p36 deletion syndrome is predicted by strain-dependent defects in Ski−/− mice Nat Genet 30 106 109 11731796
FitzPatrick DR Carr IM McLaren L Leek JP Wightman P 2003 Identification of SATB2 as the cleft palate gene on 2q32-q33 Hum Mol Genet 12 2491 2501 12915443
Murray JC Daack-Hirsch S Buetow KH Munger R Espina L 1997 Clinical and epidemiologic studies of cleft lip and palate in the Philippines Cleft Palate Craniofac J 34 7 10 9003905
Romitti PA Munger RG Murray JC Daack-Hirsch S Hanson JW 1998 The effect of follow-up on limiting participation bias in genetic epidemiologic investigations Eur J Epidemiol 14 129 138 9556171
Wilkie AOM Tang Z Elanko N Walsh S Twigg SRF 2000 Functional haploinsufficiency of the human homeobox gene MSX2 causes defects in skull ossification Nat Genet 24 387 390 10742103
Castanet M Park SM Smith A Bost M Léger J 2002 A novel loss-of-function mutation in TTF-2 is associated with congenital hypothyroidsm, thyroid agenesis and cleft palate Hum Mol Genet 11 2051 2059 12165566
Debeer P de Ravel TJL Devriendt K Fryns JP Huysmans C 2002 Human homologues of Osr1 and Osr2 are not involved in a syndrome with distal limb deficiencies, oral abnormalities, and renal defects Am J Hum Genet 111 455 456
Zhang HT Fei QY Chen F Qi QY Zou W 2003 Mutational analysis of the transforming growth factor β receptor type I gene in primary non-small cell lung cancer Lung Cancer 40 281 287 12781426
Nickerson DA Tobe VO Taylor SL 1997 PolyPhred: Automating the detection and genotyping of single nucleotide substitutions using fluorescence-based resequencing Nucleic Acids Res 25 2745 2751 9207020
Cann HM Toma C Cazes L Legrand MF Morel V 2002 A human diversity cell line panel Science 296 261 262 11954565
Blencowe BJ 2000 Exonic splicing enhancers: Mechanism of action, diversity and role in human genetic diseases Trends Biochem Sci 25 106 110 10694877
Graveley BR 2000 Sorting out the complexity of SR protein functions RNA 6 1197 1211 10999598
Wang Z Rolish ME Yeo G Tung V Mawson M 2004 Systematic identification and analysis of exonic splicing silencers Cell 119 831 845 15607979
Sunyaev S Ramensky V Bork P 2000 Towards a structural basis of human non-synonymous single nucleotide polymorphisms Trends Genet 16 198 200 10782110
Sunyaev S Ramensky V Koch I Lathe W 3rd Kondrashov AS 2001 Prediction of deleterious human alleles Hum Mol Genet 10 591 597 11230178
Ramesnky V Bork P Sunyaev S 2002 Human non-synonymous SNPs: Server and survey Nucleic Acids Res 30 3894 3900 12202775
Ranade K Chang MS Ting CT Pei D Hsiao CF 2001 High-throughput genotyping with single nucleotide polymorphisms Genome Res 11 1262 1268 11435409
Shi M Caprau D Dagle J Christiansen L Christensen K 2004 Application of kinetic polymarase chain reaction and molecular beacon assays to pooled analyses and high-throughput genotyping for candidate genes Birth Defects Res A Clin Mol Teratol 70 65 74 14991913
Laird NM Horvath S Xu X 2000 Implementing a unified approach to family-based tests of association Genet Epidemiol 19 Suppl 1 S36 42
Rabinowitz D Laird N 2000 A unified approach to adjusting association tests for population admixture with arbitrary pedigree structure and arbitrary missing marker information Hum Hered 50 211 223 10782012
Horvath S Xu X Laird NM 2001 The family based association test method: strategies for studying general genotype-phenotype associations Eur J Hum Genet 9 301 306 11313775
Bonferroni CE 1936 Teoria statistica delle classi e calcolo delle probabilità Pubblicazioni del R Istituto Superiore di Scienze Economiche e Commerciali di Firenze 8 3 62
Vieira AR Karras JC Orioli IM Castilla EE Murray JC 2002 Genetic origins in a South American clefting population Clin Genet 62 458 463 12485193
Vieira AR Orioli IM Castilla EE Cooper ME Marazita ML 2003 MSX1 and TGFB3 contribute to clefting in South America J Dent Res 82 289 292 12651933
Weinberg CR 1999 Allowing for missing parents in genetic studies of case-parent triads Am J Hum Genet 64 1186 1193 10090904
De Felice M Ovitt C Biffali E Rodriguez-Mallon A Arra C 1998 A mouse model for hereditary thyroid dysgenesis and cleft palate Nat Genet 19 395 398 9697704
Jiang R Lan Y Chapman HD Shawber C Norton CR 1998 Defects in limb, craniofacial, and thymic development in Jagged2 mutant mice Genes Dev 12 1046 1057 9531541
Zhao Y Guo YJ Tomac AC Taylor NR Grinberg A 1999 Isolated cleft palate in mice with a targeted mutation of the LIM homeobox gene Lhx8 Proc Natl Acad Sci U S A 96 15002 15006 10611327
Karolyi J Erickson RP Liu S Killewald L 1990 Major effect on teratogen-induced facial clefting in mice determined by a single genetic region Genetics 126 201 205 2227380
Lammer EJ Shaw GM Iovannisci DM van Waes J Finnell RH 2004 Maternal smoking and the risk of orofacial clefts: Susceptibility with NAT1 and NAT2 polymorphisms Epidemiology 15 150 156 15127906
Lan Y Ovitt CE Cho ES Maltby KM Wang Q 2004 Odd-skipped related 2 (Osr2) encodes a key intrinsic regulator of secondary palate growth and morphogenesis Development 131 3207 3216 15175245
Celli J Duijf P Hamel BCJ Bamshad M Kramer B 1999 Heterozygous germline mutations in the p53 homolog p63 are the cause of EEC syndrome Cell 99 143 153 10535733
Peters H Neubuser A Kratochwil K Balling R 1998 Pax9-deficient mice lack pharyngeal pouch derivatives and teeth and exhibit craniofacial and limb abnormalities Genes Dev 12 2735 2747 9732271
Reich A Sapir A Shilo B 1999 Sprouty is a general inhibitor of receptor tyrosine kinase signaling Development 126 4139 4147 10457022
Jerome LA Papaioannou VE 2001 DiGeorge syndrome phenotype in mice mutant for the T-box gene, Tbx1 Nat Genet 27 286 291 11242110
Yagi H Furutani Y Hamada H Sasaki T Asakawa S 2003 Role of TBX1 in human del22q11.2 syndrome Lancet 362 1366 1373 14585638
Bush JO Lan Y Jiang R 2004 The cleft lip and palate defects in Dancer mutant mice result from gain of function of the Tbx10 gene Proc Natl Acad Sci U S A 101 7022 7027 15118109
Kaartinen V Voncken JW Shuler C Warburton D Bu D 1995 Abnormal lung development and cleft palate in mice lacking TGF-beta-3 indicates defects of epithelial-mesenchymal interaction Nat Genet 11 415 421 7493022
Proetzel G Pawlowski SA Wiles MV Yin M Boivin GP 1995 Transforming growth factor-B3 is required for secondary palate fusion Nat Genet 11 409 414 7493021
|
16327884
|
PMC1298935
|
CC BY
|
2021-01-05 07:59:37
|
no
|
PLoS Genet. 2005 Dec 2; 1(6):e64
|
utf-8
|
PLoS Genet
| 2,005 |
10.1371/journal.pgen.0010064
|
oa_comm
|
==== Front
PLoS GenetPLoS GenetpgenplgeplosgenPLoS Genetics1553-73901553-7404Public Library of Science San Francisco, USA 1632788510.1371/journal.pgen.0010071plge-01-05-13Research ArticleSpecies Choice for Comparative Genomics: Being Greedy Works Species Choice for Comparative GenomicsPardi Fabio *Goldman Nick EMBL-European Bioinformatics Institute, Wellcome Trust Genome Campus, Hinxton, United KingdomKruglyak Leonid EditorPrinceton University, United States of America* To whom correspondence should be addressed. E-mail: [email protected] 2005 2 12 2005 27 10 2005 1 6 e7123 8 2005 27 10 2005 Copyright: © 2005 Pardi and Goldman.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.Several projects investigating genetic function and evolution through sequencing and comparison of multiple genomes are now underway. These projects consume many resources, and appropriate planning should be devoted to choosing which species to sequence, potentially involving cooperation among different sequencing centres. A widely discussed criterion for species choice is the maximisation of evolutionary divergence. Our mathematical formalization of this problem surprisingly shows that the best long-term cooperative strategy coincides with the seemingly short-term “greedy” strategy of always choosing the next best single species. Other criteria influencing species choice, such as medical relevance or sequencing costs, can also be accommodated in our approach, suggesting our results' broad relevance in scientific policy decisions.
Synopsis
What would happen if sequencing centres around the world were to choose genomes without consulting each other and without devising long-term strategies? When several parties are involved in decisions with interacting consequences, experience teaches that cooperation and planning are usually necessary to guarantee the best result. Similarly, in computer science, “greedy” algorithms—which construct solutions by iteratively taking the best immediate choice—are rarely the best option to solve a problem. The authors show, however, that in the context of choosing species for comparative genomics, cooperation and planning can be kept to a minimum without affecting the quality of the global result: a greedy algorithm applied to the problem of maximising the evolutionary divergence among species chosen from a known phylogeny is proven to guarantee optimal solutions.
Citation:Pardi F, Goldman N (2005) Species choice for comparative genomics: Being greedy works. PLoS Genet 1(6): e71.
==== Body
Introduction
Comparing biological sequences has enormous potential for increasing our knowledge about their function, structure, and evolution, an idea that has been applied virtually everywhere in computational biology. Comparative studies are now performed on a genomic scale, requiring the sequencing of entire genomes [1,2] or significant parts of them [3]. Choosing the right species for sequencing is therefore crucial. This involves two distinct decisions: first a range of species over which comparisons will be made is identified, and then a number of them are selected for actual sequencing. The first decision specifies what is known as the phylogenetic scope [4] or lineal scope [5] and is made largely on the basis of the biology the species are required to share. Different research communities are focusing on different scopes—for example, yeasts [6], nematodes [7], fruit flies [8], mammals [9], and primates [10]—corresponding to the investigation of functional elements of different biological importance.
In this article, we deal with the second decision: selecting the genomes to sequence from the chosen scope. Although this decision is determined by a variety of factors [11], chief among them is the objective of maximising the evolutionary divergence among the chosen species: the more diverse the genomes being compared, the more we can observe the different paths taken by evolution and learn about the features common to all species in the phylogenetic scope. Maximising evolutionary divergence has, for example, been advocated as a way to attain maximum sensitivity in the detection of conserved genomic regions [3,12]—regions that accumulate substitutions at a rate significantly lower than the genome-wide average. These regions are likely to be functional, as the simplest explanation for this phenomenon is the action of purifying selection (for example, see [1,3,10]), and the characterisation of non-coding conserved regions is of particular interest because their function remains unclear [9,13]. Although a maximally divergent set of species does not necessarily guarantee maximum statistical power for detecting evolutionary conservation [5], it is probably advantageous for all practical phylogenetic scopes: counterexamples are likely to arise only for (evolutionarily) very wide phylogenetic scopes, which are unrealistic in practice due to the resulting difficulty of alignment [12] and the pooling of species with different biologies.
Formalizing the problem of selecting species to maximise divergence is straightforward. Consider a phylogenetic tree connecting all the species in the chosen scope, with branch lengths representing the amount of molecular evolution between nodes in the tree. The divergence of a set of species is defined as the total branch length of the subtree connecting them (Figure 1A). The problem then becomes: given that we have already sequenced some species, and now have resources to sequence k additional species, which should we choose in order to maximise the divergence of the resulting set?
Figure 1 Phylogenetic Scopes and Divergence of Sets of Species
(A) Phylogenetic scope comprising hypothetical species A, B, C, D, and E. Numbers are branch lengths indicating evolutionary distances (not necessarily reflecting temporal distances). The subtree connecting species B, C, and E is shown in red and has divergence 1 + 3 + 1 + 5 + 2 + 4 = 16. Applying the greedy algorithm always produces maximally divergent extensions of the original set. For example, the subsets constructed starting with B—BE (divergence 11), BCE (16), BCDE (19)—have maximum divergence among those obtainable by adding one, two, and three additional species, respectively. The series AE (12), ACE (17), ACDE (20) is optimal among all possible subsets of two, three, and four species.
(B) Phylogenetic scope comprising placental mammals that have been or are being sequenced (in red) and candidates for future sequencing (derived from [17]). If five groups choose the next five targets for sequencing using the greedy strategy described in the text, the following species (in blue) will be selected (in order): (1) tenrec, (2) hedgehog, (3) rock hyrax, (4) tree shrew, (5) dog-faced fruit bat (a megabat). Within the phylogenetic scope shown, this is guaranteed to be the choice of five species that maximises the total resulting divergence. These species have recently been announced amongst targets for future sequencing [9].
In what follows, we give a simple algorithm which we prove solves this problem. We also consider, and answer, the novel question of whether different sequencing actors (groups, institutes, consortia) need to cooperate when choosing genomes: does lack of coordination and planning lead to “suboptimal” choices of genomes? While this paper assumes that optimality coincides with maximum divergence, as defined above, our results also hold for many more general species choice criteria (see Materials and Methods for details).
Results/Discussion
Imagine adopting the following “greedy” algorithm for the divergence maximisation problem: repeatedly select one species that adds the most divergence to the previously chosen ones, until k species have been added. A greedy strategy might be suspected of “short-sightedness,” i.e., leading to suboptimal solutions. We can imagine realising that a better solution could have been devised if we had considered the problem of choosing all k species at once. Perhaps surprisingly, this cannot happen. Whatever alternative strategy we devise, no better solution than that provided by the greedy algorithm is obtained. This proposition is exemplified in Figure 1A and formally proven in Materials and Methods. Note that even when the set of species previously sequenced was not optimal, the greedy algorithm guarantees the best possible subsequent extension.
Greedy algorithms are well known in computer science and often fail to guarantee optimal solutions [14]. Our result is not only of algorithmic interest, but has consequences for real-life strategies for genome sequencing. Figure 1B shows an imaginary scenario (perhaps not too far from reality) in which the genomes of a number of placental mammals have already been sequenced, and others are candidates for future sequencing. Imagine that a number of groups each have the resources to sequence one more mammal. How should they behave in order to ensure that a maximally divergent set of species is obtained? Is some sort of cooperation necessary?
Clearly, openness regarding each group's decision is necessary, since if one decides to sequence, say, the cat, the others must avoid sequencing this or any other closely related feline. Similarly, within the framework of maximising divergence, the real-life choice to sequence the rat [2] just after the mouse [1] was far from optimal. But apart from communicating their intentions, is real cooperation among the groups necessary? Applying the result described above, it is apparent that the answer is no. If every group selfishly (“greedily”) decides to sequence the genome that at the moment of choice is the most “appealing”—i.e., adds the most divergence to the set of species already sequenced or previously chosen by the other groups—then the best possible outcome is guaranteed. Another practical consequence of the optimality of the greedy algorithm is that no planning is needed, either. Specifically, no consideration of next (or any future) year's resources is necessary when determining priorities for this year's expenditure.
The greedy algorithm also guarantees an optimal solution even when other criteria for evaluating species' importance—not only divergence—are taken into account: for example, proximity to a particularly interesting species (such as human), genome size, knowledge of the species' biology, or amenability to laboratory research [11] (see Materials and Methods for further discussion). Because of this flexibility, the optimality of the greedy strategy also applies in choosing species for purposes outside comparative genomics: clearly, for genome sequencing tout-court (even when comparison is not the first use of the genome sequence) and, interestingly, for biodiversity conservation [15,16], where divergence maximisation is also considered an important objective.
If genome (or conservation) scientists follow a seemingly short-term strategy—involving neither planning nor cooperation in the choice of future genomes for sequencing (or species for conservation)—then, provided they are open about their choices, they are guaranteed the best long-term strategy.
Materials and Methods
Correctness of the greedy algorithm.
A result related to ours has been independently obtained by Steel [16], whose study concentrated on its relevance in biodiversity conservation. Steel proves that the application of the greedy algorithm on a maximally divergent set of species always results in other, larger, maximally divergent sets of species. Here, we additionally prove that applying the greedy algorithm to an initial set that is not maximally divergent results in optimal extensions of the initial set.
The idea of the proof is the following. We first prove (Theorem 1; see below) that applying a greedy choice to further extend an already optimally extended set of species always results in another optimally extended set of species. Since the first step of the greedy algorithm necessarily results in an optimally extended set, subsequent steps will construct only other optimally extended sets (Corollary 1; see below). The greedy algorithm can therefore be used to construct optimal extensions of any desired size.
Notation.
TS is a tree connecting the species in set S (coinciding with its leaves). Branches in TS are assumed to have non-negative lengths. Letters I, X, and Y will always denote subsets of S; k is a non-negative integer.
Definitions.
The tree spanning X, denoted by TX, is the smallest subtree of TS connecting all the species in X. A path is a sequence of adjacent branches in TS. The terminal path of TX leading to x (in X), is the path from TX−{x} to x. The divergence of X, denoted by δ(X), is the sum of all the branch lengths in TX. Y is a k-extension of X if Y can be obtained by adding to X k species not in X. X is a maximally divergent k-extension (k-MDE) of I if (a) X is a k-extension of I, and (b) for every k-extension Y of I, δ(Y) ≤ δ(X). We call a 1-MDE of X a greedy extension of X and denote it by X+. Note that X+ need not be unique, but any X+ will satisfy the theorem below. We will also say that X+ is obtained from X through a greedy step.
We now prove that the application of a greedy step to a maximally divergent extension (X) of an initial set (I) necessarily results in another maximally divergent extension (X+). Informally, we show that however any extension (Y) with the same size as X+ is constructed, a set that is at least as divergent as Y can be obtained from X by adding one species in Y to X. Therefore the greedy step, which can add any species to X—not only those in Y—will necessarily lead to a total divergence in X+ that is at least as great as that in Y. X+ therefore has maximum divergence among all its equally sized extensions of the initial set.
Figure 2 Representative Example for the Scenario Described in the Proof of Theorem 1.
TX is depicted in blue and TY-{x} in green. Species (leaves) in Y are represented by filled circles.
Theorem 1.
Consider sets I and X, where X is a k-MDE of I, and 2 ≤ |X| < |S|. Then X+ is a (k + 1)-MDE of I.
Proof. Let Y be any (k + 1)-extension of I. By the lemma below, there exists at least one terminal path of TY, leading to a leaf x not in X (and therefore not in I), which is completely contained in the path from TX to x (see Figure 2). Then
as X is a k-MDE of I, and
length of the path from TY−{x} to x ≤ length of the path from TX to x,
as the second path contains the first. Thus,
by summing the terms above. But
by the definition of X+. Therefore,
Since the last inequality holds for any (k + 1)-extension Y of I, X+ is a (k + 1)-MDE of I.
Observation.
Theorem 1 claims that the greedy extension of any k-MDE is a (k + 1)-MDE, assuming that the k-MDE has at least two species. This assumption ensures that either I is nonempty or k ≠ 1. In fact, if we have both I empty and k = 1, the theorem is not true: in this case, any 1-extension X of the empty set has δ(X) = 0 and is maximal. However, not every X+ will be maximal.
Corollary 1.
Let I be non-empty. The iterated application of any number k of greedy steps to I (i.e., the greedy algorithm) results in a k-MDE of I.
Proof. By induction: one greedy step results in the 1-MDE of I; if h ≥ 1 greedy steps construct an h-MDE of I, then by Theorem 1 one more step will construct an (h + 1)-MDE of I.
Corollary 2.
Let X be a maximally divergent set of h species (with h ≥ 2). Applying the greedy algorithm to X for k steps results in a maximally divergent set of h + k species.
Proof. Apply Theorem 1 with I empty, and observe that k-MDEs of the empty set are maximally divergent sets of k species. It should be noted that Corollary 2 has been proven directly by Steel [16].
Lemma.
Suppose 2 ≤ |X| < |Y|. Then there exists a leaf x in Y − X such that the path from TX to x completely contains the terminal path of TY leading to x.
Proof. Suppose the contrary. Then, for all x in Y − X, either (A) TX is contained in a subtree of TS that departs from the terminal path of TY leading to x, or (B) TX overlaps with the terminal path of TY leading to x (see Figure 3).
Figure 3 Representative Examples for Two Scenarios in the Proof of the Lemma
In both examples, TX is in blue and TY-{x} in green, and species (leaves) in Y are represented by filled circles. The two scenarios in the proof of the lemma, A and B, are illustrated correspondingly in (A) and (B), respectively.
Both (A) and (B) imply the presence of one or more leaves of X in one of the subtrees of TS that depart from the terminal path of TY leading to x. Clearly, none of these leaves can be in Y. There is at least one of these leaves (an element of X − Y) for each terminal path of TY leading to a species x not in X. Since |Y| > 2, all of these terminal paths are distinct; therefore, there are exactly |Y − X| of them and at least one leaf in X − Y for each of them, i.e., we have |X − Y| ≥ |Y − X|. But this is equivalent to |X| ≥ |Y|, which contradicts the lemma's assumptions.
Example.
For the particular case |X| = 2 and |Y| = 3, it is easy to see that the lemma holds by looking at all six possible topologically distinct cases, depicted in Figure 4.
Figure 4 Topologically Distinct Phylogenetic Trees for Two Sets of Species, X and Y, such that |X| = 2 and |Y| = 3.
X and TX are depicted in dark blue, leaves in Y are denoted with circles, and a possible choice for x (satisfying the requirements in the lemma), with the path from TX to x, in light blue.
Divergence maximisation can formalise other criteria for species selection.
Evolutionary divergence is not the only criterion guiding the selection of species for sequencing [11]. The perfect example of this comes from the decision to sequence both the mouse [1] and the rat [2], which are evolutionarily relatively close. These species were chosen because they are very well known model organisms, well suited for experimental studies, and medically relevant. It is important to note that preference towards the selection of particular species—for whatever reason—can also be formalised using a divergence maximisation approach. If we extend the terminal branch leading to each species by an amount reflecting that species' estimated importance, then application of our greedy algorithm to this modified tree leads to an optimal compromise between maximising “real” evolutionary divergence and including “preferred” species.
What kind of criteria may one take into account? The mouse-rat example already suggests some of these: deep knowledge of an organism's biology should be an advantage, as should its suitability for experimental (genetic) studies. Furthermore, we might have an intrinsic interest in one particular organism in the phylogenetic scope, and therefore we will tend to select species that are closely related to it, as these will probably share many of the genetic features we are interested in. The typical example of this is the human, but in almost every phylogenetic scope a “pivotal” species can be identified, usually a traditional model organism. The pivotal species need not be extant: one could be interested in an extinct organism, for example in reconstructing ancestral sequences or genome structure [18]. Scientific reasons are not the only ones playing a role; as in every human activity, economic interests have a crucial impact, and we expect many plant and animal genomes to be selected for sequencing on the basis of potential applications in biotechnology. Finally, one should not underestimate the importance of sequencing costs, which clearly favour species with small genome sizes.
Once these criteria are somehow quantified—which is easy at least for sequencing costs or evolutionary proximity to a pivotal species—and some idea of their relative importance defined, then we can calculate for each species a “preference score” proportional to the weighted average of that species' scores under the various criteria. We can then extend each species' terminal branch by its preference score. In practice, it may not be possible to quantify these criteria or relative weights in a generally accepted manner. Nevertheless, we can imagine that some tree modified in this way could account for the evaluation of what is “appealing” in reality being influenced by more than simply evolutionary divergence. Then greedy behaviour of sequencing groups—always choosing the currently most “appealing” species—coincides with the greedy algorithm applied to this tree, and our result provides reassurance that such behaviour will lead to an optimal solution with respect to real-life evaluations.
Note that here we assumed that it is possible to formalise the sequencing “value” of a set of species in the way described above, i.e., as the divergence of a suitably constructed tree. This is not true for all conceivable criteria for evaluating species sets, but is true at least for those that can be represented as per-lineage additive measures of value. We believe that most real-life criteria for choice [11] fall into this category.
We thank Mike Steel for helpful discussion at the Mathematics of Evolution and Phylogeny Conference (2005) in Paris and for pointing out the possibility of applying the greedy algorithm to more general criteria of species importance. FP is a member of St. Catharine's College, University of Cambridge. This work was supported by the European Molecular Biology Laboratory and by a Wellcome Trust fellowship to NG.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. FP and NG conceived the study and wrote the paper. FP derived the proof of correctness of the greedy algorithm.
A previous version of this article appeared as an Early Online Release on October 27, 2005 (DOI: 10.1371/journal.pgen.0010071.eor).
==== Refs
References
International Mouse Genome Sequencing Consortium 2002 Initial sequencing and comparative analysis of the mouse genome Nature 420 520 561 12466850
Rat Genome Sequencing Consortium 2004 Genome sequence of the brown Norway rat yields insights into mammalian evolution Nature 428 493 521 15057822
Thomas JW Touchman JW Blakesley RW Bouffard GG Beckstrom-Sternberg SM 2003 Comparative analyses of multi-species sequences from targeted genomic regions Nature 424 788 793 12917688
Cooper GM Brudno M Green ED Batzoglou S Sidow A 2003 Quantitative estimates of sequence divergence for comparative analyses of mammalian genomes Genome Res 13 813 820 12727901
McAuliffe JD Jordan MI Pachter L 2005 Subtree power analysis and species selection for comparative genomics Proc Natl Acad Sci U S A 102 7900 7905 15911755
Kellis M Patterson N Endrizzi M Birren B Lander ES 2003 Sequencing and comparison of yeast species to identify genes and regulatory elements Nature 423 241 254 12748633
Stein LD Bao ZR Blasiar D Blumenthal T Brent MR 2003 The genome sequence of Caenorhabditis briggsae: A platform for comparative genomics PLoS Biol 1 e5 12929205
Richards S Liu Y Bettencourt BR Hradecky P Letovsky S 2005 Comparative genome sequencing of Drosophila pseudoobscura: Chromosomal, gene, and cis-element evolution Genome Res 15 1 18 15632085
Margulies EH Vinson JP Miller W Jaffe DB Lindblad-Toh K 2005 An initial strategy for the systematic identification of functional elements in the human genome by low-redundancy comparative sequencing Proc Natl Acad Sci U S A 102 4795 4800 15778292
Boffelli D McAuliffe J Ovcharenko D Lewis KD Ovcharenko I 2003 Phylogenetic shadowing of primate sequences to find functional regions of the human genome Science 299 1391 1394 12610304
O'Brien SJ Eizirik E Murphy WJ 2001 Genomics—On choosing mammalian genomes for sequencing Science 292 2264 2266 11423643
Eddy SR 2005 A model of the statistical power of comparative genome sequence analysis PLoS Biol 3 e10 15660152
Bejerano G Siepel A Kent WJ Haussler D 2005 Computational screening of conserved genomic DNA in search of functional noncoding elements Nature Methods 2 535 545 16170870
Cormen TH Leiserson CE Rivest RL Stein C 2001 Introduction to algorithms. Chapter 16. 2nd Edition Cambridge (Massachusetts) MIT Press 370 404
Nee S May RM 1997 Extinction and the loss of evolutionary history Science 278 692 694 9381180
Steel M 2005 Phylogenetic diversity and the greedy algorithm Syst Biol 54 527 529 16051588
Madsen O Scally M Douady CJ Kao DJ DeBry RW 2001 Parallel adaptive radiations in two major clades of placental mammals Nature 409 610 614 11214318
Blanchette M Green ED Miller W Haussler D 2004 Reconstructing large regions of an ancestral mammalian genome in silico Genome Res 14 2412 2423 15574820
|
16327885
|
PMC1298936
|
CC BY
|
2021-01-05 07:59:37
|
no
|
PLoS Genet. 2005 Dec 2; 1(6):e71
|
utf-8
|
PLoS Genet
| 2,005 |
10.1371/journal.pgen.0010071
|
oa_comm
|
==== Front
PLoS PathogPLoS PathogppatplpaplospathPLoS Pathogens1553-73661553-7374Public Library of Science San Francisco, USA 1633339510.1371/journal.ppat.001003505-PLPA-RA-0087R3plpa-01-04-02Research ArticleEubacteriaA Novel Endogenous Inhibitor of the Secreted Streptococcal NAD-Glycohydrolase Immunity Factor for SPN (IFS)Meehl Michael A 1Pinkner Jerome S 1Anderson Patricia J 23Hultgren Scott J 1Caparon Michael G 1*
1 Department of Molecular Microbiology, Washington University School of Medicine, St. Louis, Missouri, United States of America
2 Department of Medicine, Washington University School of Medicine, St. Louis, Missouri, United States of America
3 Howard Hughes Medical Institute, Washington University School of Medicine, St. Louis, Missouri, United States of America
Ghosh Partho EditorUniversity of California at San Diego, United States of America* To whom correspondence should be addressed. E-mail: [email protected] 2005 2 12 2005 1 4 e3526 7 2005 24 8 2005 Copyright: © 2005 Meehl et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.The Streptococcus pyogenes NAD-glycohydrolase (SPN) is a toxic enzyme that is introduced into infected host cells by the cytolysin-mediated translocation pathway. However, how S. pyogenes protects itself from the self-toxicity of SPN had been unknown. In this report, we describe immunity factor for SPN (IFS), a novel endogenous inhibitor that is essential for SPN expression. A small protein of 161 amino acids, IFS is localized in the bacterial cytoplasmic compartment. IFS forms a stable complex with SPN at a 1:1 molar ratio and inhibits SPN's NAD-glycohydrolase activity by acting as a competitive inhibitor of its β-NAD+ substrate. Mutational studies revealed that the gene for IFS is essential for viability in those S. pyogenes strains that express an NAD-glycohydrolase activity. However, numerous strains contain a truncated allele of ifs that is linked to an NAD-glycohydrolase−deficient variant allele of spn. Of practical concern, IFS allowed the normally toxic SPN to be produced in the heterologous host Escherichia coli to facilitate its purification. To our knowledge, IFS is the first molecularly characterized endogenous inhibitor of a bacterial β-NAD+−consuming toxin and may contribute protective functions in the streptococci to afford SPN-mediated pathogenesis.
Synopsis
The gram-positive bacterium Streptococcus pyogenes is a human pathogen that causes a wide range of infections from pharyngitis (“strep throat”) to invasive necrotizing fasciitis (“flesh-eating disease”). While strep throat responds to antibiotic therapy, more invasive infections caused by S. pyogenes often require surgical intervention. It is currently unknown exactly how the bacteria can switch between the different types of infection, but one possibility is via a mechanism by which the bacterium injects a bacterial protein toxin (S. pyogenes NAD-glycohydrolase [SPN]) into human skin cells, causing their death. In this study, the authors have shown that the injected toxin also has the ability to affect the bacteria. A second protein neutralizes SPN to ensure the bacteria are immune to its toxic effects. Consequently, S. pyogenes has developed a valuable weapon in its arsenal to promote its survival by ensuring the safe production of SPN, through its own protection by immunity factor for SPN, enabling the delivery of active SPN into human cells. The process reported in this paper may ultimately help create therapeutic inhibitors of SPN and possibly other SPN-like toxins implicated in microbial disease progression.
Citation:Meehl MA, Pinkner JS, Anderson PJ, Hultgren SJ, Caparon MG (2005) A novel endogenous inhibitor of the secreted streptococcal NAD-glycohydrolase. PLoS Pathog 1(4): e35.
==== Body
Introduction
Bacterial pathogens secrete a multitude of factors that are utilized to advance the infectious process. Many of the secreted factors exhibit an enzymatic activity that is directed against host-specific targets or are activated by host-specific functions. However, a few secreted enzymes are quite promiscuous and have the ability to adversely affect both the microbe and the host cell. Because of this potential self-toxicity, bacteria must develop mechanisms to protect themselves from the deleterious effects of these universally toxic enzymes in order to successfully use them in pathogenesis. One toxic enzyme, the secreted nicotinamide adenine dinucleotide (NAD)–glycohydrolase of Streptococcus pyogenes (SPN, also named NGA [1]), has recently been shown to be injected into the host cell cytoplasm via a specialized translocation process known as cytolysin-mediated translocation (CMT) [2,3]. However, how S. pyogenes manages the potential self-toxicity of SPN is unknown.
SPN is one of several secreted toxins that are thought to contribute to the pathogenesis of the numerous diseases that S. pyogenes can cause. These range from superficial (pharyngitis, impetigo) to life threatening (toxic shock syndrome, necrotizing fasciitis) [4]. The contribution that any one toxin makes to a specific S. pyogenes disease is generally not understood. However, SPN has several activities that suggest it may be important for pathogenesis. As an NAD-glycohydrolase, its most well characterized activity is its ability to cleave β-NAD+ at the ribose-nicotinamide bond to generate ADP-ribose and the potent vasoactive compound nicotinamide [5−7]. Similar to several other NAD-glycohydrolases, SPN has also been reported to have a cyclase activity capable of converting β-NAD+ into cyclic ADP-ribose, a potent second messenger for calcium mobilization [8]. The observation that SPN can transfer ADP-ribose to certain synthetic substrates has suggested that SPN may ADP-ribosylate an important host protein in order to modify the function of that protein [1]. However, the roles that any of these activities may contribute to pathogenesis remains to be established.
Studies using in vitro models of streptococcal pathogenesis have provided evidence that SPN can alter host cell behavior following its translocation into the cytosolic compartment [2,3]. One effect of intracellular SPN is an enhanced cytotoxic response that results in the rapid death of the infected host cell [2,3]. The basis of the cytotoxic response is not understood; however, any of SPN's enzymatic activities could potentially have deleterious effects on host cell viability. For example, if left unchecked, SPN's robust NAD-glycohydrolase activity would likely result in a significant reduction in the intracellular stores of β-NAD+. As a co-factor, β-NAD+ is universally important in numerous essential redox and energy-producing biological reactions (for a recent review, see [9]). Depletion of β-NAD+ would be likely be toxic for both the intoxicated host cell and the streptococcal cell producing SPN. Toxicity for bacterial cells would explain the fact that it has not been possible to clone and express the intact gene for SPN in heterologous hosts such as Escherichia coli (M. G. Caparon, unpublished data). Thus, successful production of SPN by its native S. pyogenes host likely requires some mechanism for management of its toxic properties.
In this report, we describe immunity factor for SPN (IFS), a novel endogenous inhibitor of the NAD-glycohydrolase activity of SPN. These studies show that the ability of S. pyogenes to express and secrete an active SPN has an essential dependence on the production of IFS in its cytoplasmic compartment. Furthermore, co-expression of SPN and IFS allows the normally toxic SPN to be produced at high levels in E. coli and allows SPN to be exported as an active enzyme into the periplasmic space. Fractionation of both recombinant and streptococcal extracts identified a cytoplasmic inhibition of NAD-glycohydrolase activity that was dependent on IFS. The mechanism of inhibition involves the ability of IFS to bind tightly to SPN and to act as a competitive inhibitor of its β-NAD+ substrate. Taken together, these data reveal how S. pyogenes protects itself from the toxic effects of SPN through its expression of IFS, a novel inhibitor of a β-NAD+−consuming microbial toxin.
Results
Efficient Cloning and Expression of SPN
The fact that it has not been possible to clone SPN in any E. coli expression vector led us to hypothesize that S. pyogenes must have a factor that protects it from any potential toxicity associated with SPN. In searching for this putative factor, we noted that the gene for SPN is present in an operon that includes the gene encoding SLO, the pore-forming component of the CMT injection pathway (Figure 1). Of interest, the operon also includes an open reading frame (spy0166,
Figure 1) that could potentially encode a small protein (161 amino acids, 18.8 kDa) and would be negatively charged at neutral pH (predicted pI = 5.39). The Spy0166 protein lacks a gram-positive Sec-dependent signal peptide (see Materials and Methods) and would therefore likely serve its role as a soluble cytoplasmic protein since the Sec pathway is the only available route for protein export in S. pyogenes. These properties are also characteristic of some protease inhibitors [10,11] and chaperones of proteins translocated by the type III secretion systems of gram-negative bacteria [12], both of which can increase the efficiency of expression of their partner proteins in E. coli and in their native hosts [13−16]. Based on these observations, we tested whether spy0166 from S. pyogenes strain JRS4 could support the expression of spn in E. coli when both genes were placed under the control of the arabinose-inducible promoter on the vector pBAD/gIIIB (see Materials and Methods). In this construct (pMAM3.18), the native signal sequence of SPN was replaced with the gene III signal sequence (for efficient periplasmic targeting) while 6X-HIS and c-myc epitope tags were grafted on the carboxy-terminal end of Spy0166. Unlike any previous SPN expression construct, pMAM3.18 efficiently transformed E. coli (>104 transformants/μg DNA) with high fidelity (no nucleotide substitutions, as determined by sequence analysis of several clones). Furthermore, there was no loss of viability following induction of the vector's promoter and fractionation of the resulting cultures revealed SPN in the periplasmic fraction and Spy0166 in the cytoplasmic fraction as detected by Western blot analysis (data not shown).
Figure 1 The spn-slo Operon Includes the Open Reading Frame spy0166
The figure depicts the spn-slo chromosomal operon from S. pyogenes M1 strain SF370 [18]. A promoter upstream of spn (bent arrow) drives expression of spn and slo. Assignment of open reading frames is taken from the annotation of the SF370 genome.
Spy0166 Inhibits the Glycohydrolase Activity of SPN
Since SPN's β-NAD+−consuming glycohydrolase activity may have contributed to the protein's apparent toxicity for E. coli, it was of interest to determine whether Spy0166 could inhibit SPN's ability to cleave β-NAD+. The strategy was to mix together periplasmic and cytoplasmic fractions in various combinations following arabinose induction of E. coli strains containing either pMAM3.18 or the vector alone. Upon analysis, the periplasmic fractions derived from pMAM3.18, but not vector-containing strains, demonstrated a robust NAD-glycohydrolase activity (compare bars 1 and 2, Figure 2). Furthermore, the activity produced by pMAM3.18 was not affected by mixing with the cytoplasmic fraction derived from a strain containing the vector alone (bar 3, Figure 2). In contrast, mixing an SPN-containing periplasmic fraction with an Spy0166-containing cytoplasmic fraction resulted in an inhibition of NAD-glycohydrolase activity to near background levels (bar 4, Figure 2), and the degree of inhibition was dependent upon the amount of Spy0166-containing cytoplasmic extract mixed with SPN (Figure 3). In the absence of SPN, Spy1066 had no glycohydrolase activity of its own (bars 5, 6, and 7, Figure 2). Based on its ability to inhibit NAD-glycohydrolase activity and support the viability of SPN-expressing E. coli, Spy0166 was renamed IFS.
Figure 2 SPN Activity Is Inhibited by IFS (Spy0166)
The labeled bars indicate the NAD-glycohydrolase activities of mixtures of isolated periplasm (as a source of SPN) and cytoplasm (as a source of Spy0166) from various E. coli strains containing the plasmids indicated. The presence (+) or absence (−) of either SPN or Spy0166 in a particular periplasmic or cytoplasmic extract is indicated. Vector is pBAD/gIIIB containing no inserted streptococcal DNA. Asterisk indicates that the level of NAD-glycohydrolase activity was below the limit of detection of 50 pmol/min. Due to its ability to inhibit NAD-glycohydrolase activity, Spy0166 was renamed IFS. Data represent the mean ± standard error of the mean derived from three independent experiments.
Figure 3 Dose-Dependent Inhibition by IFS
NAD-glycohydrolase activity produced by a constant concentration of periplasmic extract (as a source of SPN) mixed with increasing concentrations of a cytoplasmic extract (as a source of IFS), both of which were prepared from TOP10(pMAM3.18). Asterisk indicates that the level of NAD-glycohydrolase activity was below the limit of detection of 50 pmol/min. Data represent the mean ± standard error of the mean derived from three independent experiments.
Two Alleles of spn and ifs
Comparison of spn operon sequences among the genomes available for several S. pyogenes strains along with sequences obtained from two other commonly studied strains (JRS4 [M6 serotype, 17] and HSC5 [M14 serotype, unpublished data]) revealed the presence of at least two distinct alleles of ifs. A distinguishing characteristic is the presence of a nonsense mutation that converts the codon encoding leucine at position 24 (TAT) to a stop codon (TAA) (Figure 4). As a result, the annotation of the genomes of SF370 [18] and MGAS8232 [19] lists the codon for methionine at position 44 as the initiation codon of the IFS open reading frame (Figure 4). The strain used to characterize IFS in the experiments described in the previous section contains the longer allele (JRS4, Figure 4). In addition, it had been previously reported that there are at least two distinct alleles of spn that differ by five nonsynonymous substitutions [1]. There was a correlation between which spn allele a given strain contained and whether or not it produced immunoreactive SPN [1]. Of the strains analyzed in this study, JRS4 is known to produce active SPN [2] and has the allele associated with active expression (JRS4-WT, Figure 5). Furthermore, the NAD-glycohydrolase activity produced by this strain is completely dependent upon SPN (SPN−, Figure 5). In contrast, HSC5 has the allele that was associated with the absence of an ability to produce a detectable SPN protein similar to that reported for M1 strain SF370 [1]. As predicted by this comparison, HSC5 failed to produce any detectable NAD-glycohydrolase activity and any immunoreactive SPN protein (compare HSC5-WT and SF370, Figure 5). One difference between JRS4, SF370, and HSC5 is that the latter two strains are strong producers of the SpeB cysteine protease, while the former is not. Since it has been reported that SpeB can completely degrade SPN [20], HSC5\′s ability to produce SPN was examined in the presence of the cysteine protease inhibitor E64. With this treatment, HSC5 now contained the SPN protein at levels similar to JRS4 (compare JRS4-WT and HSC5-E64, Figure 5). However, the HSC5 SPN still lacked any detectable NAD-glycohydrolase activity (HSC5-E64, Figure 5). This was not a result of E64 treatment or any other inhibitory factor in HSC5 supernatants, because (1) E64 did not inhibit JRS4 SPN (data not shown), (2) elimination of protease activity through mutation of SpeB's active site residue or a deletion of an essential activator of speB transcription allowed HSC5 to produce SPN that still exhibited no detectable NAD-glycohydrolase activity (HSC5-C192S and HSC5-RopB−, Figure 5), and (3) mixing HSC5 and JSR4 supernatants did not result in inhibition of the JRS4 SPN (data not shown). Consistent with a previous report [20], strain SF370 produced a low amount of SPN in the presence of E64 by Western blot analysis with no detectable NAD-glycohydrolase activity (data not shown). Taken together, these data indicate that in addition to possessing two alleles for ifs, the streptococcal population contains at least two major alleles of spn, and the product of one of these lacks detectable NAD-glycohydrolase activity.
Figure 4 Two Alleles of ifs
A multiple alignment of ifs from several S. pyogenes strains is shown. A nonsense mutation in the codon for leucine 24 (indicated by asterisk) produces a truncated ifs open reading frame in several strains. The circles above the sequence show the position of several other polymorphic residues. The ifs loci were taken from the following genomes: M1 (SF370, ifs = Spy0166), M13 (MGAS315, ifs = Spy_M30129), M18 (MGAS8232, ifs = Spy_M180164), and strains JRS4 and HSC5.
Figure 5 Expression of Enzymatically Active and Inactive SPN Proteins
A Western blot analysis of cell-free culture supernatant fluids from various S. pyogenes strains is shown. The blot was developed with an antiserum that recognizes the proteins indicated on the left, including SPN, full-length SLO, and a truncated form of SLO (SLO*) generated through a specific cleavage by the streptococcal SpeB cysteine protease [52]. The lanes labeled under JRS4 include JRS4 itself (WT), SPN mutant SPN1 (SPN−), and SLO mutant SLO1 (SLO−), and under SF370, SF370 itself (WT). Lanes under HSC5 include HSC5 itself (WT), HSC5 grown in the presence of the protease inhibitor E64 (E64), and mutants of HSC5: catalytically deficient SpeB mutant JWR10 (C192S) and protease regulatory mutant MNN100 (RopB−). The NAD-glycohydrolase activity titer (NADase titer) of each supernatant fluid is shown at the bottom of each lane. The data shown are representative of results from three independent experiments.
Cytoplasmic Extracts of S. pyogenes Contain an Allele-Dependent Inhibitory Activity
The data presented above suggest that the function of IFS is to inhibit the NAD-glycohydrolase activity of SPN in the cytoplasmic compartment of S. pyogenes. In addition, available data indicate linkage between spn and ifs alleles. Strains that contain the spn allele associated with activity (e.g., JRS4) contain the larger ifs open reading frame [see also MGAS315 genome, 21], while those that possess the spn allele that does not express activity (e.g., HSC5, SF370) have truncated ifs [see also MGAS8232 genome, 19]. To directly test whether S. pyogenes contains an inhibitory activity for SPN and the possible influence of the ifs allele, the ability of cytoplasmic extracts from various strains to inhibit NAD-glycohydrolase activity was evaluated. Consistent with the behavior of IFS cloned from JRS4 (see above), cytoplasmic extracts of JRS4 contained an activity that inhibited the NAD-glycohydrolase activity of the secreted form of JRS4 SPN (Figure 6). In contrast, cytoplasmic extracts from either HSC5 or SF370 did not contain an activity that inhibited the NAD-glycohydrolase activity of JRS4 SPN (Figure 6). Extracts prepared from HSC5 lacked activity even when prepared in the presence of the protease inhibitor E64 (data not shown).
Figure 6 A Cytoplasmic Inhibitory Activity
The NAD-glycohydrolase activities of a mixture of JRS4 supernatant (as a source of SPN) and cytoplasmic extracts prepared from the indicated strains are shown. Asterisk indicates that the level of NAD-glycohydrolase activity was below the limit of detection of 50 pmol/min. Data represent the mean ± standard error of the mean derived from three independent experiments.
Cytoplasmic Inhibitory Activity Was Dependent on IFS
To determine if the inhibitory activity detected in JRS4 was due to IFS, an attempt was made to construct an in-frame deletion in the gene for IFS. The method for mutagenesis proceeds via the generation of a tandem duplication of mutant and wild-type alleles in the genome that can resolve to either allele by recombination. Typically, chromosomes with either allele are isolated at similar frequencies among the progeny [22]. However, while it was possible to produce the merodiploid intermediate strain at the ifs locus, all progeny recovered following resolution of the duplication contained a copy of the wild-type allele (30 of 30 tested). In contrast, when mutagenesis was conducted in a JRS4 SPN− mutant, progeny with the mutant or wild-type ifs alleles were recovered at equal frequencies (four mutants, four wild-type, eight tested). These data suggest that IFS is essential for viability when SPN is encoded in the genome. Consistent with this, when progeny of the SPN− mutant were analyzed, only those progeny with the wild-type ifs allele could inhibit SPN NAD-glycohydrolase activity (compare IFS+ Rev. to IFS−, Figure 7). Inhibitory activity was restored in the IFS− mutant upon the introduction of an HA-tagged version of the JRS4 ifs on a plasmid (compare pIFS-HA to Vector, Figure 7). The presence of IFS in the cytoplasm of the complementing strain was verified by Western blot analysis using antiserum to the HA epitope tag (data not shown). Taken together, these data indicate that in the absence of IFS, expression of SPN is likely lethal for JRS4 and that the inhibitory activity detected in cytoplasmic extracts is dependent on IFS.
Figure 7 Inhibitory Activity Is Due to IFS
The NAD-glycohydrolase activities of a mixture of JRS4 supernatant (as a source of SPN) and cytoplasmic extracts prepared from several derivatives of JRS4 SPN− mutant SPN1 are shown. Lanes include SPN− mutant itself (Parent), a derivative of SPN− mutant with an in-frame deletion in ifs (IFS−), a sibling of the SPN− IFS− deletion mutant that reverted to wild-type ifs (IFS+ Rev.), the SPN− IFS− deletion mutant containing a plasmid-encoded HA-tagged version of IFS (IFS− pIFS-HA), and the pABG5 plasmid vector lacking ifs (IFS− Vector). Asterisk indicates that the level of NAD-glycohydrolase activity was below the limit of detection of 50 pmol/min. Data represent the mean ± standard error of the mean derived from three independent experiments.
An SPN-IFS Complex
The ability of IFS to inhibit SPN's enzymatic activity suggested that the two proteins interact. To test this, pull-down assays were conducted following mixing periplasmic extracts (as a source of SPN) and cytoplasmic extracts (as a source of IFS) prepared from various E. coli strains. In the first assay, the cytoplasmic extract was prepared from a strain expressing IFS with C-terminal 6X-His and c-myc epitope tags (IFS-H) and periplasm was prepared from a strain expressing SPN and IFS. Following mixing, IFS-H was isolated via selective binding to Ni-NTA agarose and any co-isolation of SPN assessed by Western blot analysis. In the absence of IFS-H, SPN demonstrated some affinity for Ni-NTA agarose under these conditions (lane 3, Figure 8A). However, in the presence of IFS-H, SPN was readily co-isolated (lane 4, Figure 8A) at levels substantially greater than background observed with cytoplasmic extracts from strains containing the vector alone (lane 5, Figure 8A) or an irrelevant 6X-His–tagged protein (lane 6, Figure 8A). The reciprocal experiment yielded similar results, where IFS (without a 6X-His tag) was selectively co-isolated only when incubated with periplasm containing 6X-His–tagged SPN (SPN-H) (lane 4, Figure 8B) and not in the absence of any periplasmic extract (lane 3, Figure 8B), the presence of periplasmic extracts prepared from strains expressing the plasmid vector alone (lane 5, Figure 8B), or an irrelevant 6X-His–tagged protein (lane 6, Figure 8B).
Figure 8 SPN and IFS Form a Complex
Ni-NTA pull-down assays detect the formation of a complex between IFS and SPN.
(A) Periplasmic extracts (periplasm) from E. coli strain TOP10(pMAM3.8) expressing an non−6X-His–tagged version of SPN (SPN), the pBAD/pIIIB vector (Vec.), or buffer (None) were mixed with cytoplasmic extracts (cytoplasm) from TOP10(pMAM3.19) that expressed a 6X-His and c-myc–tagged version of IFS (IFS-H), the pBAD/gIIIB vector (Vec.), a plasmid expressing a 6X-His and c-myc–tagged calmodulin (Cal-H), or buffer (None). Proteins released following pull-down with Ni-NTA agarose were analyzed by Western blot with an antiserum to detect SPN, as shown.
(B) Periplasmic extracts from TOP10(pMAM3.18) expressing a strain a 6X-His–tagged version of SPN (SPN-H) were mixed with cytoplasmic extracts prepared from TOP10(pMAM3.21), which expresses IFS with a c-myc epitope, but lacking a 6X-His tag. Other designations are as in (A). Proteins released following pull-down with Ni-NTA agarose were analyzed by Western blot with an antiserum to detect the c-myc tags of IFS and Cal-H, as shown. Controls are purified IFS and SPN that were not subjected to pull-down, and the migration of IFS, SPN, and Cal-H is indicated to the right.
Characteristics of the SPN-IFS Complex
The ability of IFS and SPN to form a complex was also assessed by size exclusion chromatography. Analysis of purified recombinant SPN (see Materials and Methods) by this method indicated an apparent molecular weight in solution of 52.4 ± 1.5 kDa (Figure 9), consistent with the predicted monomer size of the mature protein based on its primary sequence (48.4 kDa) and several previous reports indicating that SPN purified from streptococcal supernatant fluids is a monomer in solution [7,23,24]. However, a similar analysis of purified recombinant IFS (see Materials and Methods) indicated an apparent molecular weight of 31.5 ± 3.2 kDa (Figure 9), a value that was unexpected based on a molecular weight of 22.0 kDa predicted from the IFS-myc-HIS primary sequence. This suggests that IFS exists as a monomer with an extended shape or as an atypical dimer with Stoke's radius typical of a 31-kDa globular protein. Next, an assembled SPN-IFS complex was purified from the cytoplasm of recombinant E. coli (TOP10 strain harboring pMAM3.18) and subjected to chromatography with the resulting size of 67.6 ± 3.2 kDa (Figure 9) that was larger that either SPN or IFS alone. Importantly, this complex eluted in a resolving fraction that followed the void volume, which was calculated at 78.6 kDa (see Materials and Methods). Assembly of complex in vitro from purified IFS and SPN resulted in a major peak eluting in the void volume, indicating a possible aggregation event occurred in vitro (data curve not shown). The size of the more physiologically relevant in vivo assembled complex (67 kDa) indicates that one SPN molecule (48 kDa) interacts with one IFS molecule (22 kDa). This was confirmed by an SDS-PAGE analysis of several fractions obtained from the leading edge of the high molecular peak that revealed the presence of both SPN and IFS (Figure 10). Analysis of SDS-PAGE gels by densitometry with comparison to known concentrations of purified SPN and IFS indicated that the molar ratio of SPN to IFS was 1:1 in these complexes, in agreement with the gel filtration data as stated above.
Figure 9 Gel Filtration Chromatography of SPN, IFS, and the SPN-IFS Complex
Purified SPN (B), IFS (C), or the SPN-IFS (A) complex was subjected to gel filtration chromatography over Superdex 75 HR 10/30, and their elution profiles are overlaid. Based on the elution of several standards (identities of the standards are shown, and their elution volumes are indicated by the corresponding arrows), the calculated molecular weights of SPN, IFS, and the complex are 52.4 ± 1.5, 31.5 ± 3.2, and 67.6 ± 3.2 kDa, respectively. These calculations were based on the average elution volumes derived from at least three independent column applications.
Figure 10 Analysis of the SPN-IFS Complex
Several fractions of the high-molecular-weight peak arising from gel filtration chromatography of an SPN-IFS complex were analyzed by SDS-PAGE as shown. Purified SPN and IFS are included for comparison. Densitometric analyses (see Materials and Methods) of similar gels containing known concentrations of purified SPN and IFS revealed a 1:1 molar stoichiometry of SPN to IFS in the complex.
IFS Is a Competitive Inhibitor of SPN
In order to determine the inhibition mechanism, the effects of IFS on the rate of hydrolysis of β-NAD+ by SPN were studied as a function of IFS concentration (Figure 11). The rate of β-NAD+ hydrolysis in the absence of IFS showed a hyperbolic increase with an apparent Km for β-NAD+ of 379 ± 74 μM and an observed Vmax of 9.9 ± 1.2 μM/min. Inhibition of SPN activity by IFS was determined by varying the concentration of IFS in the reactions. Increasing concentrations of IFS decreased the rate of β-NAD+ hydrolysis. Simultaneous fitting of these curves by a competitive model yielded a Km,app of 348 ± 54 μM for β-NAD+, a Vmax,obs of 9.6 ± 0.8 μM/min, and a KI,app of 2.0 ± 0.3 nM (Figure 11). The observed Km,app and Vmax,obs values were consistent to those obtained by fitting of the data by the Michaelis-Menten equation in the absence of any inhibitor. Simultaneous fitting of the curves by uncompetitive and noncompetitive models yielded parameters that fit the data poorly (fits not shown) and were inconsistent to the parameters obtained by fits of β-NAD+ hydrolysis in the absence of any inhibitor. The results demonstrate that SPN was inhibited by IFS in a competitive manner as indicated by the consistent Michaelis constants. The results also indicated that the complex between SPN and IFS was tight in comparison with the substrate β-NAD+.
Figure 11 Kinetic Analysis of the Inhibition Mechanism of SPN by IFS
The observed initial rate (vobs) of proteolysis of β-NAD+ by SPN (0.125 U/ml) was assessed in the absence (•) and presence of varying concentrations of IFS (○, 0.47 nM; ▴, 0.93 nM; Δ, 1.3 nM; □, 1.9 nM; ▪, 2.3 nM). The lines represent the nonlinear least-squares analysis of the competitive inhibition model with the parameters described in the text. Initial rates were measured, and the data were analyzed as described in Materials and Methods.
Discussion
In this study, we have described IFS, a novel endogenous inhibitor of SPN, the secreted NAD-glycohydrolase of S. pyogenes. IFS forms a stable complex with active SPN and functions as a competitive inhibitor of its β-NAD+ substrate. Furthermore, IFS proved to be essential for the ability of S. pyogenes to express SPN. As a practical consequence, co-expression of IFS and SPN proved to be a highly successful strategy for production of SPN in E. coli. To our knowledge, ifs represents the first molecularly characterized gene product encoding an endogenous inhibitor for a β-NAD+−consuming microbial toxin or for any NAD-glycohydrolase to date from either the prokaryotic or eukaryotic kingdoms.
In order to produce and secrete universally toxic molecules, bacterial pathogens have evolved several strategies to protect themselves from the self-toxicity of these molecules. For example, some toxins, like the calmodulin-dependent adenylate cyclases or the cholesterol-dependent cytolysins, require a co-factor found exclusively in a host compartment for activity (for reviews, see [25,26]). Other toxic enzymes, most notably broad-spectrum proteases, are secreted as inactive precursors whose conversion to an active form can be subsequently regulated both temporally and spatially [27]. A less common strategy involves the production of a cytoplasmic inhibitor. Although an inhibitor of the streptococcal SpeB cysteine protease was recently discovered [28], the best-studied example of this class is the staphostatin-staphopain system of the staphylococci [29]. Staphopains are secreted cysteine proteases that are specifically inhibited by the staphostatins, which are small (approximately 13 kDa) and acidic proteins that, similar to IFS, reside in the staphylococcal cytoplasm. Also like IFS, the staphostatins form a stable noncovalent complex at a 1:1 molar ratio with their cognate staphopain in their fully folded and active conformations [30]. Three-dimensional structures of the staphostatin-staphopain complex have revealed a competitive mechanism of inhibition whereby the staphostatin binds to the staphopain in a substrate-like manner forming a long-lived inhibitor-enzyme complex [30−34]. The observations that co-expression of a staphostatin increases the efficiency of staphopain production when expressed in E. coli [13] and that deletion of a staphostatin has a profound affect on the ability of the mutant to produce its cell wall and export proteins [35] have suggested that the function of the staphostatins is to prevent improper degradation of cytoplasmic proteins by premature activation of staphopains in the cytoplasmic compartment.
Similar to a staphostatin, the function of IFS may prevent the premature activation of the enzyme in the bacterial cytosolic compartment. However, while numerous inhibitors for various proteases have been described, including a recent discovery of an SpeB endogenous inhibitor [28], endogenous inhibitors of β-NAD+−consuming toxins are less common. Several prior reports have characterized endogenous NAD-glycohydrolase inhibitory activity in extracts of Bacillus subtilis and Mycobacterium phlei extracts [36,37]. However, identification of the genes that encode these activities has not been reported. The ability of IFS to bind in a tight complex with a 1:1 molar stoichiometry with SPN and to act as a competitive inhibitor of SPN's β-NAD+ substrate implies that IFS may function similarly to the staphostatins and form a complex that makes contact with residues at or near the catalytic center of SPN. Alternatively, IFS may act as an effective molecular mimic of β-NAD+ and function as a noncleavable substrate. Should the latter mechanism prove to be the case, IFS may act as a broad-spectrum inhibitor of β-NAD+−consuming toxins.
An important and unusual feature of both IFS and the staphostatins is that they are located in the cytoplasmic compartment, yet they bind and inhibit the activities of their cognate enzymes when these are in their active and fully folded conformations. While this property is valuable if the primary function of these proteins to inhibit the potentially toxic activities of the prematurely activated toxins, it also implies that these proteins become folded prior to their secretion. In this scenario, IFS may act to protect the cell from a small population of SPN molecules that drift off the Sec secretion pathway and then fold in the cytoplasmic compartment. This would explain why IFS is essential in both native and heterologous hosts that contain NAD-glycohydrolase−proficient SPN. This function would also be consistent with the apparent linkage between the NAD-glycohydrolase−deficient spn allele and the truncated allele for ifs. These strains produced an SPN polypeptide but lacked a cytoplasmic inhibitory activity. Thus, in the absence of enzymatic activity, an IFS-mediated inhibitory activity is not required, and this would likely reduce any selective pressure to maintain the full-length ifs allele. The loss of inhibitory activity does not rule out the possibility that IFS may have some other essential role. Multiple functions have been attributed to the chaperones of the type III secretion pathway, including secretion targeting of the effector-chaperone complex, unfolding cognate effectors, protection of effectors from proteases, and establishing a temporal hierarchy for effector secretion [38]. Many of these same functions are likely required for successful targeting for Sec pathway secretion and delivery of SPN to the CMT pathway. Thus, any of these functions may also be provided by IFS. However, it remains to be determined whether the truncated allele is expressed and whether it is required for expression of the NAD-glycohydrolase−deficient SPN or its translocation into the host cell cytosol via CMT.
The existence of an NAD-glycohydrolase−deficient allele of spn raises some interesting questions on the function of SPN and its repertory of enzymatic activities in pathogenesis. An important role is supported by the fact that several studies have shown that the gene for SPN is present in virtually all S. pyogenes isolates examined to date [1,39]. Two major alleles of spn were found, and there was an associated lack of immunoreactive protein for one allele [1]. We have shown here that the lack of detectable protein is directly caused by an SpeB-dependent degradation event through the use of a biochemical inhibitor and mutagenesis of speB (see Figure 5), and this observation was also noted in a recent study [20]. It is unclear if the variability in detectable SPN protein is due to an alteration in SpeB expression or whether SpeB is inefficient at degrading the active SPN allele. Furthermore, we have demonstrated an association of the spn allele exhibiting NAD-glycohydrolase activity with a full-length ifs allele. Due to the small sampling of strains used in this study, a more thorough study on the correlation of these alleles could be useful to establish whether the ifs nonsense mutation induces a resulting mutation in spn to prevent self-toxicity by β-NAD+ depletion. Nevertheless, the fact that the NAD-glycohydrolase−deficient allele is apparently widely distributed in the streptococcal population indicates it may have an as-yet-unidentified contribution to pathogenesis that is independent of an NAD-glycohydrolase activity. Thus, a more thorough understanding of SPN's activities and its contribution to virulence via the CMT pathway will require considerably more experimentation.
Materials and Methods
Bacterial strains and culture conditions.
Molecular cloning experiments utilized E. coli TOP10 (Invitrogen, Carlsbad, California, United States). The S. pyogenes strains JRS4, SLO1 (JRS4 SLO−), SPN1 (JRS4 SPN−), SF370, HSC5, JWR10 (HSC5 speBC192S), and MNN100 (HSC5 RopB−) have previously described [2,40−42]. Luria-Bertaini medium was used for culture of E. coli, while routine culture of S. pyogenes utilized Todd-Hewitt medium (BBL) supplemented with 0.2% yeast extract (Difco, BD Biosciences, San Diego, California, United States) (THY medium). For certain assays (see below), S. pyogenes was cultured in Dulbecco's modified Eagle media (without glucose, phenol red, or pyruvate) supplemented with 2.5% yeast extract and 2% glucose (DMEM-YE medium). Antibiotics were routinely used to maintain selection for plasmids and were added to media at the following concentrations: kanamycin, 50 μg/ml for E. coli and 500 μg/ml for S. pyogenes; erythromycin, 750 μg/ml for E. coli and 1 μg/ml for S. pyogenes; and ampicillin, 100 μg/ml for E. coli. Where indicated, the cysteine protease inhibitor E-64 (Sigma, St. Louis, Missouri, United States) was added to THY medium at a final concentration of 28 μM.
DNA and computational techniques.
Plasmid DNA was isolated by standard techniques and used to transform chemically competent E. coli [43] and to transform S. pyogenes by electroporation [44]. Restriction endonucleases, ligases, and polymerases were used according to the manufacturers' recommendations. Chromosomal DNA was purified from S. pyogenes as previously described [44]. Fidelity of all DNA sequences generated by PCR was verified using fluorescently labeled dideoxynucleotides (Big Dye terminators; Applied Biosystems, Foster City, California, United States) and the appropriate oligonucleotide primers in DNA sequencing reactions according to the recommendations of the manufacturer. Identification of gram-positive secretion signal peptides utilized the SignalP model [45]. The alignment of spy0166 (ifs) open-reading frames was generated using the ClustalW algorithm [46].
Cloning and expression of spn and ifs.
Plasmids for the cloning and expression of spn and ifs in E. coli were based on the pBAD/gIIIB expression vector (Invitrogen). Primers MAM98 and MAM84 (Table S1) were designed to amplify a single fragment containing spn and ifs from JRS4 chromosomal DNA so that the sequences encoding the signal sequence of spn were replaced with a sequence encoding a 6X-His affinity tag. The fragment was digested with NcoI and BstBI using sites embedded in the primers and inserted between the NcoI and BstBI sites of pBAD/gIIIB. In the resulting plasmid (pMAM3.14), spn was grafted in-frame with a gene III signal sequence followed by the 6X-His sequence and ifs was grafted in-frame with vector sequences specifying a C-terminal c-myc epitope tag followed by a 6X-His affinity tag. A derivative of pMAM3.14 was constructed with primers MAM101 and MAM102 (Table S1) using an “inside-out” PCR strategy [47] in order to remove the 6X-His segment introduced into ifs. The resulting plasmid (pMAM3.18) contains an additional PstI site introduced by the primers that was used to recircularize the PCR product. A similar strategy using primers MAM93B and MAM84 (Table S1) and a pMAM3.14 template was used to construct a plasmid (pMAM3.8) that lacked the segment encoding the 6X-His tag upstream of spn, while retaining the 6X-His tag sequence on ifs. Primers MAM103 and MAM84 (Table S1) were used to amplify and insert ifs between the NcoI and BstB1 sites of pBAD/gIIIB. The resulting plasmid (pMAM3.19) expressed IFS with C-terminal c-myc and 6X-His tags in the cytoplasmic compartment of its E. coli host. This plasmid was modified to remove the sequence encoding the 6X-His tag with primers MAM101 and MAM102 using an “inside-out” strategy as described above. The resulting plasmid (pMAM3.21) expresses IFS with a c-myc epitope tag.
Construction of ifs in-frame deletion.
Primers MAM78B and MAM79B (Table S1) amplified a 1.4-kb fragment containing the entire JRS4 ifs open reading frame with flanking sequences in spn and slo. The fragment was inserted into a standard vector (pCRII; Invitrogen) using XhoI sites embedded in the primers. An “inside-out” PCR strategy using primers MAM80 and MAM81 (Table S1) introduced an in-frame deletion of 0.46 kb into the ifs open-reading frame. Sites for SalI (embedded in primers MAM80 and MAM81) were used to insert the fragment containing the deletion and flanking sequences into the shuttle vector pJRS233 [48]. The resulting plasmid (pMAM3.3) was then used in attempts to replace the wild-type ifs allele as described in the text according to a previously described method [42]. Wild-type and mutant alleles were distinguished by PCR amplification of chromosomal DNA using primers MAM79B and MAM87S (Table S1) that yielded products of 1,083 bp (wild-type) or 627 bp (ifs deletion), and assignment was confirmed by PCR using primers MAM110 and MAM112 (Table S1) that yielded a 480-bp product only in the presence of wild-type ifs.
Ectopic expression of ifs in S. pyogenes.
Primers MAM108 and MAM112 (Table S1) were used to amplify ifs with its native ribosome-binding site from JRS4 chromosomal DNA with the addition of DNA encoding a C-terminal HA epitope tag. The resulting 0.51-kb fragment was inserted between the EcoRI and PstI of the E. coli/streptococcal shuttle vector pABG5 [49] using sites embedded in the primers. In the resulting plasmid (pMAM3.32), expression of the gene for the modified IFS (IFS-HA) is controlled by the rofA promoter [49]. Expression of IFS-HA following introduction of pMAM3.32 into S. pyogenes was confirmed in Western blot analysis of cell extracts (see below) using an anti-HA antiserum (Sigma).
Assays for SPN.
Expression of SPN protein was evaluated by Western blot analyses of S. pyogenes culture supernatant fluids [2] or from E. coli periplasmic or cytoplasmic fractions [50] using an antiserum that cross-reacts with SPN (anti-SLO antibody, lot No. 7317011; Golden West Biologicals, Temecula, California, United States) [2]. The NAD-glycohydrolase activity of SPN was assessed by either endpoint titer or determination of units of activity (in picomoles of β-NAD+ cleaved per minute), both as previously described [2]. For the latter, more than 50 units represents the lower limit of detection of the assay.
Assays for IFS.
Expression of SPN protein was evaluated by Western blot analyses of cytoplasmic extracts of S. pyogenes or from periplasmic or cytoplasmic fractions [50] of E. coli using antisera that recognize the appropriate HA (see above) or c-myc (Sigma) epitope tags. To isolate streptococcal cytoplasmic extracts, bacterial cells (washed twice and resuspended in 10 mM Tris [pH 8.0]) were lysed using a high-speed reciprocating shaking device (FP-120; Savant Instruments, Holbrook, New York, United States), and the extract was separated by centrifugation (15,000×g, 10 min, 4 °C). The success of streptococcal cell lysis was verified by visualization of cytoplasmic proteins by SDS-PAGE analysis and supported by the gain of inhibitory activity in the complemented ifs− mutant strain. Activity assays for the ability of IFS to inhibit NAD-glycohydrolase activity were as follows. For E. coli fractions, equal volumes of induced periplasm (0.05 μg total protein), cytoplasm (0.5 μg total protein), or buffer (1× PBS) were mixed and incubated with β-NAD+ (0.133 μM) for 0 to 50 min in a 96-well microtiter plate. For S. pyogenes fractions, cell-free overnight culture supernatants from JRS4 (diluted 1:15 in PBS) were mixed with streptococcal cytoplasmic extracts (50 μg total protein) or buffer (1× PBS) and incubated with β-NAD+ as described above. Units of NAD-glycohydrolase activity detectable at the end of the incubation period were then determined as described previously [2].
Ni-NTA agarose pull-down assay.
The various E. coli strains described in the text were cultured under inducing conditions (see below), and cytoplasmic and periplasmic fractions were prepared [50]. For pull-down assays, periplasm (7.5 μg total protein), cytoplasm (75 μg total protein), or buffer (1× PBS) was coincubated for 15 min at room temperature prior to the addition of 50 μl of a 50% solution of Ni-NTA agarose (Qiagen, Valencia, California, United States). The mixtures were incubated for an additional 15 min at room temperature, and the agarose beads were recovered by centrifugation. Following aspiration of the supernatant fluids, the beads were washed three times in an equal volume of wash buffer (50 mM NaH2PO4, 300 mM NaCl, 20 mM imidazole [pH 8.0]), and bound proteins were eluted by incubation with an equal volume of elution buffer (50 mM NaH2PO4, 300 mM NaCl, 250 mM imidazole [pH 8.0]) for 15 min at room temperature. Beads were removed by centrifugation, and proteins in the supernatant fluids were subjected to Western blot analyses using antisera to detect SPN or IFS. Controls included extracts prepared from E. coli strains containing the vector alone (pBAD/gIIIB) or pBAD/gIII/calmodulin (pCALM), which expresses calmodulin with 6X-His and c-myc tags (Invitrogen).
Purification of SPN.
TOP10 cells containing pMAM3.18 were induced at mid-log phase (OD600 0.5) for 4 h using 0.2% l-arabinose (final concentration). Periplasm was isolated as previously described [50] and dialyzed against buffer composed of 50 mM phosphate buffer and 300 mM NaCl (pH 7.0). The sample was then subjected to metal-affinity chromatography over a TALON Superflow resin using the washing and elution conditions recommended by the manufacturer (BD Biosciences). Fractions containing SPN were dialyzed in a buffer containing 20 mM MES (pH 5.8) and subjected to chromatography over a Resource S ion-exchange column (Amersham Biosciences, Little Chalfont, United Kingdom) using a linear NaCl gradient (0 to 500 mM) to elute SPN. Fractions were analyzed by SDS-PAGE, and those fractions containing greater than 95% SPN were pooled (see Figure 10).
Purification of IFS.
TOP10 cells containing pMAM3.19 were induced at mid-log phase (OD600 0.5) for 4 h using 0.2% l-arabinose (final concentration). Cells were prepared by extraction of periplasm and then resuspended in TALON column buffer (300 mM NaCl, 50 mM Na-phosphate [pH 7.0]) and lysed by sonication. The lysate was then subjected to metal affinity chromatography over a TALON Superflow resin using conditions recommended by the manufacturer (BD Biosciences). Fractions containing IFS were dialyzed against buffer containing 20 mM Tris-HCl (pH 8.5) and subjected to chromatography over a Resource Q ion exchange column (Amersham Biosciences) using a linear NaCl gradient (0 to 500 mM) to elute IFS. Fractions were analyzed by SDS-PAGE, and those fractions containing greater than 95% SPN were pooled (see Figure 10).
Molecular weight estimates.
Size exclusion chromatography was used to estimate the molecular weights of SPN, IFS, and the SPN-IFS complex. All chromatography was performed on a Superdex 75 HR 10/30 column (Amersham Biosciences) using a flow rate of 0.3 ml/min. Low-molecular-weight protein standards (Amersham Biosciences, see Figure 9) and test samples were equilibrated and developed in a buffer of 20 mM Tris (pH 8.5) plus 100 mM NaCl. Standards were analyzed (200 μl) on three separate occasions, and the average elution volumes were recorded. The value of Kav was determined for each standard using the equation Kav = (Ve − Vo)/(Vt − Vo), where Ve is the elution volume, Vo is the void volume (7.9 ml), and Vt is the bed volume of the column (24 ml). For each standard, Kav was plotted against log Mr (molecular weight). Molecular weight estimates represented the average derived from at least three independent experiments.
Stoichiometry of the SPN-IFS complex.
The protein concentration of the SPN-IFS complex isolated by size exclusion chromatography (see above) was determined using the bicinchoninic acid assay (Sigma) and a BSA protein standard (Sigma). Various concentrations of the complex were subjected to SDS-PAGE along with known concentrations of purified SPN and IFS. Gels were stained with Coomassie R-250, and an image captured on Kodak Image Station 2000MM was analyzed using Kodak 1D Image Analysis Software. Total pixel intensities of each band were obtained using the ROI (region of interest) function of the software. The intensities of bands from the known molar concentrations of SPN and IFS included on each gel were used to generate standard curves that were used to calculate the concentration of SPN and IFS from the pixel intensities of bands resolved from the complex. The calculated molar ratio reported was based on two independent experiments, each of which yielded an identical result.
Kinetic analysis of the inhibitory mechanism of IFS.
The inhibition mechanism of IFS was determined in kinetic assays of SPN activity monitored by fluorescence detection of hydrolysis of β-NAD+ (Sigma), as previously described [8]. Briefly, JRS4 cells were grown overnight in DMEM-YE and centrifuged at 6,000 rpm for 10 min, and collected supernatants were sterile-filtered. One unit of SPN activity was defined as the activity in 1 ml of supernatant from the overnight growth. At least three preparations of SPN were used throughout these studies, and they yielded consistent and overlapping rates of hydrolysis. SPN (0.125 U/ml) was incubated with varying concentrations of β-NAD+ (up to 1.2 mM) in PBS at 37 °C for varying times (up to 50 min). Reactions were quenched by the addition of 5N NaOH, and the fluorescence of residual β-NAD+ was detected using excitation wavelength of 380 nm and an emission wavelength of 455 nm on a Perkin-Elmer LS55B using the plate reader accessory. The residual concentration of β-NAD+ was calculated as previously described [2]. Plots of the concentration of β-NAD+ hydrolysis with time were fit to a straight line to obtain the observed velocity (vobs). The rates of β-NAD+ hydrolysis as a function of β-NAD+ concentration were fit by the Michaelis-Menten equation to determine the apparent Michaelis constant (Km,app) and the maximum rate (Vmax,obs) [51]. The inhibition mechanism of IFS was determined in reactions of varying concentrations of purified IFS (up to 2.3 nM) with SPN (0.125 U/ml) and varying concentrations of β-NAD+ as described above. The observed velocities as a function of β-NAD+ and varying concentrations of IFS were simultaneously fit by competitive, noncompetitive, and uncompetitive inhibition models to determine the apparent inhibition constant (KI,app), the apparent Michaelis constant (Km,app), and the observed maximal velocity (Vmax,obs) [51]. The model that gave the best fit was used in determination of the inhibition mechanism. Nonlinear least-squares analysis was performed with Scientist (Micromath, Salt Lake City, Utah, United States). Errors in the reported parameters are ±2 SDs.
Supporting Information
Table S1 Primers Used in This Study
(31 KB PDF)
Click here for additional data file.
Accession Numbers
The GenBank (http://www.ncbi.nlm.nih.gov/Genbank) accession numbers for the genes and gene products discussed in this paper are IFS (DQ093072), Spy_M30129 (NC_004070), Spy_M180164 (NC_003485), spy0166 (NC_002737), strain HSC5 (DQ093073), and strain JRS4 (DQ093072).
We are indebted to Joe Vogel for his helpful comments and valuable insights. We thank Petra Levin for her interest and suggestions, as well as Melody Neely and Jason Rosch for construction of mutant strains. This work was supported by U.S. Public Health Service grants AI064721 (MGC) and DK064540 (SJH) from the National Institutes of Health. PJA was supported by National Scientist Development Award 0530110N from the American Heart Association.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. MAM, JSP, PJA, and MGC conceived and designed the experiments. MAM and JSP performed the experiments. MAM, JSP, PJA, and MGC analyzed the data. PJA, SJH, and MGC contributed reagents/materials/analysis tools. MAM, PJA, and MGC wrote the paper.
Abbreviations
β-NAD+β-nicotinamide adenine dinucleotide
CMTcytolysin-mediated translocation
IFSimmunity factor for SPN
SPN
Streptococcus pyogenes NAD-glycohydrolase
==== Refs
References
Stevens DL Salmi DB McIndoo ER Bryant AE 2000 Molecular epidemiology of nga and NAD glycohydrolase/ADP-ribosyltransferase activity among Streptococcus pyogenes causing streptococcal toxic shock syndrome J Infect Dis 182 1117 1128 10979908
Madden JC Ruiz N Caparon M 2001 Cytolysin-mediated translocation (CMT): A functional equivalent of type III secretion in gram-positive bacteria Cell 104 143 152 11163247
Bricker AL Cywes C Ashbaugh CD Wessels MR 2002 NAD+-glycohydrolase acts as an intracellular toxin to enhance the extracellular survival of group A streptococci Mol Microbiol 44 257 269 11967084
Cunningham MW 2000 Pathogenesis of group A streptococcal infections Clin Microbiol Rev 13 470 511 10885988
Bernheimer AW Lazarides PD Wilson AT 1957 Diphosphopyridine nucleotidase as an extracellular product of streptococcal growth and its possible relationship to leukotoxicity J Exp Med 106 27 37 13439112
Carlson AS Kellner A Bernheimer AW Freeman EB 1957 A streptococcal enzyme that acts specifically upon diphosphopyridine nucleotide: Characterization of the enzyme and its separation from streptolysin O J Exp Med 106 15 26 13439111
Gerlach D Ozegowski JH Gunther E Vettermann S Kohler W 1996 Purification and some properties of streptococcal NAD-glycohydrolase FEMS Microbiol Lett 136 71 78 8919458
Karasawa T Takasawa S Yamakawa K Yonekura H Okamoto H 1995 NAD(+)-glycohydrolase from Streptococcus pyogenes shows cyclic ADP-ribose forming activity FEMS Microbiol Lett 130 201 204 7649441
Berger F Ramirez-Hernandez MH Ziegler M 2004 The new life of a centenarian: Signalling functions of NAD(P) Trends Biochem Sci 29 111 118 15003268
Massimi I Park E Rice K Muller-Esterl W Sauder D 2002 Identification of a novel maturation mechanism and restricted substrate specificity for the SspB cysteine protease of Staphylococcus aureus
J Biol Chem 277 41770 41777 12207024
Rzychon M Sabat A Kosowska K Potempa J Dubin A 2003 Staphostatins: An expanding new group of proteinase inhibitors with a unique specificity for the regulation of staphopains, Staphylococcus spp. cysteine proteinases Mol Microbiol 49 1051 1066 12890028
Wattiau P Woestyn S Cornelis GR 1996 Customized secretion chaperones in pathogenic bacteria Mol Microbiol 20 255 262 8733225
Wladyka B Puzia K Dubin A 2005 Efficient co-expression of a recombinant staphopain A and its inhibitor staphostatin A in Escherichia coli
Biochem J 385 181 187 15320867
Losada LC Hutcheson SW 2005 Type III secretion chaperones of Pseudomonas syringae protect effectors from Lon-associated degradation Mol Microbiol 55 941 953 15661015
Darwin KH Miller VL 2000 The putative invasion protein chaperone SicA acts together with InvF to activate the expression of Salmonella typhimurium virulence genes Mol Microbiol 35 949 960 10692170
Tucker SC Galan JE 2000 Complex function for SicA, a Salmonella enterica serovar typhimurium type III secretion-associated chaperone J Bacteriol 182 2262 2268 10735870
Scott JR Guenthner PC Malone LM Fischetti VA 1986 Conversion of an M- group A streptococcus to M+ by transfer of a plasmid containing an M6 gene J Exp Med 164 1641 1651 3534131
Ferretti JJ McShan WM Ajdic D Savic DJ Savic G 2001 Complete genome sequence of an M1 strain of Streptococcus pyogenes
Proc Natl Acad Sci U S A 98 4658 4663 11296296
Smoot JC Barbian KD Van Gompel JJ Smoot LM Chaussee MS 2002 Genome sequence and comparative microarray analysis of serotype M18 group A Streptococcus strains associated with acute rheumatic fever outbreaks Proc Natl Acad Sci U S A 99 4668 4673 11917108
Aziz RK Pabst MJ Jeng A Kansal R Low DE 2004 Invasive M1T1 group A Streptococcus undergoes a phase-shift in vivo to prevent proteolytic degradation of multiple virulence factors by SpeB Mol Microbiol 51 123 134 14651616
Beres SB Sylva GL Barbian KD Lei B Hoff JS 2002 Genome sequence of a serotype M3 strain of group A Streptococcus: Phage-encoded toxins, the high-virulence phenotype, and clone emergence Proc Natl Acad Sci U S A 99 10078 10083 12122206
Caparon M 2000 Genetics of group A streptococci Rood JI Gram-positive pathogens Washington (DC) ASM Press 53 65
Fehrenbach FJ 1969 Gel-filtration behaviour and molecular weight of NAD-glycohydrolase (EC 3.2.2.5) from streptococci in column chromatography on Sephadex gels J Chromatogr 41 43 52 5780864
Grushoff PS Shany S Bernheimer AW 1975 Purification and properties of streptococcal nicotinamide adenine dinucleotide glycohydrolase J Bacteriol 122 599 605 236282
Tweten RK Parker MW Johnson AE 2001 The cholesterol-dependent cytolysins Curr Top Microbiol Immunol 257 15 33 11417120
Ahuja N Kumar P Bhatnagar R 2004 The adenylate cyclase toxins Crit Rev Microbiol 30 187 196 15490970
Lyon WR Caparon MG 2003 Trigger factor-mediated prolyl isomerization influences maturation of the Streptococcus pyogenes cysteine protease J Bacteriol 185 3661 3667 12775704
Kagawa TF O'Toole P Cooney JC 2005 SpeB-Spi: A novel protease-inhibitor pair from Streptococcus pyogenes
Mol Microbiol 57 650 666 16045611
Dubin G 2005 Proteinaceous cysteine protease inhibitors Cell Mol Life Sci 62 653 669 15770418
Filipek R Rzychon M Oleksy A Gruca M Dubin A 2003 The staphostatin-staphopain complex: A forward binding inhibitor in complex with its target cysteine protease J Biol Chem 278 40959 40966 12874290
Filipek R Potempa J Bochtler M 2005 A comparison of staphostatin B with standard mechanism serine protease inhibitors J Biol Chem 280 14669 14674 15644332
Filipek R Szczepanowski R Sabat A Potempa J Bochtler M 2004 Prostaphopain B structure: A comparison of proregion-mediated and staphostatin-mediated protease inhibition Biochemistry 43 14306 14315 15518582
Dubin G Krajewski M Popowicz G Stec-Niemczyk J Bochtler M 2003 A novel class of cysteine protease inhibitors: Solution structure of staphostatin A from Staphylococcus aureus
Biochemistry 42 13449 13456 14621990
Rzychon M Filipek R Sabat A Kosowska K Dubin A 2003 Staphostatins resemble lipocalins, not cystatins in fold Protein Sci 12 2252 2256 14500882
Shaw LN Golonka E Szmyd G Foster SJ Travis J 2005 Cytoplasmic control of premature activation of a secreted protease zymogen: Deletion of staphostatin B (SspC) in Staphylococcus aureus 8325–4 yields a profound pleiotropic phenotype J Bacteriol 187 1751 1762 15716447
Everse KE Everse J Simeral LS 1980
Bacillus subtilis NADase and its specific protein inhibitor Methods Enzymol 66 137 144 6246385
Davis WB 1980 Identification of a nicotinamide adenine dinucleotide glycohydrolase and an associated inhibitor in isoniazid-susceptible and -resistant Mycobacterium phlei
Antimicrob Agents Chemother 17 663 668 6249194
Ghosh P 2004 Process of protein transport by the type III secretion system Microbiol Mol Biol Rev 68 771 795 15590783
Ajdic D McShan WM Savic DJ Gerlach D Ferretti JJ 2000 The NAD-glycohydrolase (nga) gene of Streptococcus pyogenes
FEMS Microbiol Lett 191 235 241 11024269
Hanski E Horwitz PA Caparon MG 1992 Expression of protein F, the fibronectin-binding protein of Streptococcus pyogenes JRS4, in heterologous streptococcal and enterococcal strains promotes their adherence to respiratory epithelial cells Infect Immun 60 5119 5125 1452345
Neely MN Lyon WR Runft DL Caparon M 2003 Role of RopB in growth phase expression of the SpeB cysteine protease of Streptococcus pyogenes
J Bacteriol 185 5166 5174 12923089
Ruiz N Wang B Pentland A Caparon M 1998 Streptolysin O and adherence synergistically modulate proinflammatory responses of keratinocytes to group A streptococci Mol Microbiol 27 337 346 9484889
Kushner S 1978 An improved method of transformation of Escherichia coli with ColE1-derived plasmids Nicosia S Genetic engineering New York Elsevier/North Holland Biomedical Press p. 173
Caparon MG Scott JR 1991 Genetic manipulation of pathogenic streptococci Methods Enzymol 204 556 586 1658571
Nielsen H Engelbrecht J Brunak S von Heijne G 1997 Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites Protein Eng 10 1 6
Thompson JD Higgins DG Gibson TJ 1994 CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice Nucleic Acids Res 22 4673 4680 7984417
Horton RM 1997 In vitro recombination and mutagenesis of DNA White BA PCR cloning protocols: From molecular cloning to genetic engineering Totowa (New Jersey) Humana Press 141 149
Perez-Casal J Price JA Maguin E Scott JR 1993 An M protein with a single C repeat prevents phagocytosis of Streptococcus pyogenes: Use of a temperature-sensitive shuttle vector to deliver homologous sequences to the chromosome of S. pyogenes
Mol Microbiol 8 809 819 8355608
Granok AB Parsonage D Ross RP Caparon MG 2000 The RofA binding site in Streptococcus pyogenes is utilized in multiple transcriptional pathways J Bacteriol 182 1529 1540 10692357
Slonim LN Pinkner JS Branden CI Hultgren SJ 1992 Interactive surface in the PapD chaperone cleft is conserved in pilus chaperone superfamily and essential in subunit recognition and assembly EMBO J 11 4747 4756 1361168
Segel IH 1975 Enzyme kinetics: Behavior and analysis of rapid equilibrium and steady state enzyme systems New York Wiley 992 p.
Pinkney M Kapur V Smith J Weller U Palmer M 1995 Different forms of streptolysin O produced by Streptococcus pyogenes and by Escherichia coli expressing recombinant toxin: Cleavage by streptococcal cysteine protease Infect Immun 63 2776 2779 7790099
|
16333395
|
PMC1298937
|
CC BY
|
2021-01-05 12:11:28
|
no
|
PLoS Pathog. 2005 Dec 2; 1(4):e35
|
utf-8
|
PLoS Pathog
| 2,005 |
10.1371/journal.ppat.0010035
|
oa_comm
|
==== Front
PLoS PathogPLoS PathogppatplpaplospathPLoS Pathogens1553-73661553-7374Public Library of Science San Francisco, USA 1634125410.1371/journal.ppat.001003905-PLPA-RA-0120R2plpa-01-04-01Research ArticleInfectious DiseasesMolecular Biology - Structural BiologyVirologyGenetics/Gene FunctionGenetics/Gene ExpressionVirusesAnimalsMus (Mouse)Functional and Genetic Analysis of Coronavirus Replicase-Transcriptase Proteins MHV-A59 Replicase-Transcriptase ProteinsSawicki Stanley G 1Sawicki Dorothea L 1Younker Diane 1¤aMeyer Yvonne 2Thiel Volker 2¤bStokes Helen 3Siddell Stuart G 3*
1 Department of Medical Microbiology and Immunology, Medical University of Ohio, Toledo, Ohio, United States of America
2 Institute of Virology, University of Würzburg, Würzburg, Germany
3 Department of Cellular and Molecular Medicine, University of Bristol, Bristol, United Kingdom
Andino Raul EditorUniversity of California at San Francisco, United States of America* To whom correspondence should be addressed. E-mail: [email protected]¤a Current address: CheCS-Environmental Health Systems, Houston, Texas, United States of America
¤b Current address: Research Department, Cantonal Hospital, St. Gallen, Switzerland
12 2005 9 12 2005 1 4 e395 8 2005 1 11 2005 Copyright: © 2005 Sawicki et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.The coronavirus replicase-transcriptase complex is an assembly of viral and cellular proteins that mediate the synthesis of genome and subgenome-sized mRNAs in the virus-infected cell. Here, we report a genetic and functional analysis of 19 temperature-sensitive (ts) mutants of Murine hepatitis virus MHV-A59 that are unable to synthesize viral RNA when the infection is initiated and maintained at the non-permissive temperature. Both classical and biochemical complementation analysis leads us to predict that the majority of MHV-A59 ORF1a replicase gene products (non-structural proteins nsp1–nsp11) form a single complementation group (cistron1) while the replicase gene products encoded in ORF1b (non-structural proteins nsp12–nsp16) are able to function in trans and comprise at least three, and possibly five, further complementation groups (cistrons II–VI). Also, we have identified mutations in the non-structural proteins nsp 4, nsp5, nsp10, nsp12, nsp14, and nsp16 that are responsible for the ts phenotype of eight MHV-A59 mutants, which allows us to conclude that these proteins are essential for the assembly of a functional replicase-transcriptase complex. Finally, our analysis of viral RNA synthesis in ts mutant virus-infected cells allows us to discriminate three phenotypes with regard to the inability of specific mutants to synthesize viral RNA at the non-permissive temperature. Mutant LA ts6 appeared to be defective in continuing negative-strand synthesis, mutant Alb ts16 appeared to form negative strands but these were not utilized for positive-strand RNA synthesis, and mutant Alb ts22 was defective in the elongation of both positive- and negative-strand RNA. On the basis of these results, we propose a model that describes a pathway for viral RNA synthesis in MHV-A59-infected cells. Further biochemical analysis of these mutants should allow us to identify intermediates in this pathway and elucidate the precise function(s) of the viral replicase proteins involved.
Synopsis
Coronaviruses infect both humans and animals and are associated mainly with respiratory and enteric diseases. The recent outbreak of SARS emphasizes the need to develop new strategies to control these infections. This paper focuses on the proteins involved in the replication of the coronavirus genome and the production of viral mRNAs in the host cell. These so-called replicase-transcriptase proteins are likely to make good targets for the development of anti-coronaviral drugs. The approach used here is to analyze conditional, temperature-sensitive mutants of Murine hepatitis virus that are normal at 33 °C (the permissive temperature) but are unable to replicate and transcribe viral RNAs at 39.5 °C (the restrictive temperature). By identifying the genetic changes responsible for these temperature-sensitive mutations and by analyzing the precise nature of the defect in RNA synthesis at the restrictive temperature, the authors are able to propose a model that describes a pathway for viral RNA synthesis in the infected cell. Further analysis of these mutants should allow the elucidation of the precise function(s) of the viral proteins involved.
Citation:Sawicki SG, Sawicki DL, Younker D, Meyer Y, Thiel V, et al. (2005) Functional and genetic analysis of coronavirus replicase-transcriptase proteins. PLoS Pathog 1(4): e39.
==== Body
Introduction
Coronaviruses are positive-strand, enveloped RNA viruses that infect vertebrates and are associated mainly with respiratory and enteric disease. They have long been recognized as important pathogens of livestock and companion animals, and they are a common cause of respiratory tract infections in humans [1–3]. More recently, a coronavirus has been identified as the causative agent of SARS, a form of atypical pneumonia in humans with a case fatality ratio of approximately 10% [4]. Clearly, there is an urgent need to develop new strategies to prevent or control coronavirus infections, and understanding the biology, replication, and pathogenesis of these viruses is an essential part of this process. Murine hepatitis virus, strain A59 (MHV-A59), is a group II coronavirus with a genome of approximately 31,400 nucleotides. The genomic RNA encodes the structural proteins of the virus, non-structural proteins involved in viral RNA synthesis (the nsp or replicase proteins), and proteins that are non-essential for replication in cell culture but appear to confer a selective advantage in vivo (accessory proteins) [1]. In the MHV-A59-infected cell, the expression of the replicase protein genes is mediated by translation of the genomic RNA, and the expression of the structural protein genes is mediated by the translation of a set of 3′-coterminal subgenomic mRNAs. The subgenomic mRNAs are produced by a unique mechanism that involves discontinuous transcription during negative-strand RNA synthesis [5–7]. The organization and expression of the MHV-A59 genome are illustrated in Figure 1.
Figure 1 Organization and Expression of the MHV-A59 Genome
The structural relationships of the MHV-A59 genome and sub-genomic mRNAs are shown. The virus ORFs are depicted as lightly shaded (replicase proteins), shaded (accessory proteins), and heavily shaded (structural proteins). The ORFs are defined by the genomic sequence of MHV-A59 as published by Coley et al. [45]. The hatched box represents the common 5′ leader sequence and the hatched circle represents the programmed (−1) frameshifting element. The translation products of the genome and sub-genomic mRNAs are depicted and the autoproteolytic processing of the ORF1a and ORF1a/ORF1b polyproteins into non-structural proteins nsp1 to nsp16 is shown. A number of confirmed and putative functional domains in the non-structural proteins are also indicated: 3CL, 3C-like cysteine proteinase; ExoN, exonuclease; HEL, superfamily 1 helicase; MT, S-adenosylmethionine-dependent 2′-O-methyl transferase; NeU, endoribonuclease; PL1, papain-like protease 1; PL2, papain-like protease 2; POL, RNA-dependent RNA polymerase; X, adenosine diphosphate-ribose 1′-phosphatase.
The 5′ proximal open reading frames (ORF) of MHV-A59 genomic RNA (ORF1a and ORF1b) are translated to produce two large polyproteins, pp1a and pp1ab, with calculated molecular masses of 496.6 and 802.8 kilodaltons, respectively. Translation of the larger pp1ab involves programmed (−1) ribosomal frameshifting [8]. During or after synthesis, these polypeptides are extensively processed by three virus-encoded proteinases to produce a membrane-bound replicase-transcriptase complex [9]. Cleavage of the replicase polyproteins is predicted to result in 16 end-products; nsp1–nsp11 encoded in ORF1a and nsp12–16 encoded in ORF1b [10]. These proteins have been shown, or are predicted to have multiple enzymatic functions, including papain-like proteases (nsp3), adenosine diphosphate-ribose 1′-phosphatase (nsp3), 3C-like cysteine proteinase (nsp5), RNA-dependent RNA polymerase (nsp12), superfamily 1 helicase (nsp13), exonuclease (nsp14), endoribonuclease (nsp15), and S-adenosylmethionine-dependent 2′-O-methyl transferase (nsp16) [11–20]. The crystallographic structures of SARS coronavirus nsp5 and nsp9 have been determined and are likely to be similar for MHV-A59 [21–23].
In the course of an infectious cycle, the MHV-A59 replicase-transcriptase complex amplifies the genomic RNA and synthesizes subgenomic mRNAs. Amplification of the genomic RNA involves full-length negative-strand templates, and the synthesis of subgenomic mRNA involves subgenome-length negative-strand templates [24,25]. The structures engaged in the replication and transcription of positive-strand MHV-A59 RNA have been characterized [26]. Approximately 70% of the replicating and transcribing structures that accumulate in infected cells are multi-stranded intermediates (replicative and transcriptive intermediate RNA, RI/TI RNA) and 30% are found in structures with only one or very few nascent strands (native replicative and transcriptive forms, RF/TF RNA). Although the structures engaged in negative-strand RNA synthesis have not yet been characterized, it is known that MHV negative-strand templates are unstable and turn over during viral replication [27].
The cis-acting RNA elements involved in the different phases of MHV RNA synthesis have been studied quite extensively. It has been shown that 5′- and 3′-UTR, as well as 5′-UTR-adjacent regions of the genome are required for MHV replication and transcription [28,29]. Also, studies on MHV, and other nidoviruses, have shown the critical role of the so-called transcription-regulating sequence (TRS) element in the discontinuous phase of the transcription process [7,30–33]. These data show that the stability of the leader-TRS/body-TRS duplex, which forms during the discontinuous extension phase of negative-strand template synthesis, is an important determinant of subgenomic mRNA abundance. However, it is also evident from these studies that the regions flanking the TRS elements have a profound affect on the amounts of subgenomic mRNAs that are produced. In the context of the discontinuous-extension model [5], this is explained as different degrees of “attenuation” at each of the TRS elements during negative-strand synthesis.
In contrast, there is still very little known about the structure, functions, and interactions of viral and cellular proteins in the replicase-transcriptase complex as it is engaged in different modes of RNA synthesis. As mentioned above, bioinformatic and biochemical studies have identified a number of (putative) enzymatic activities associated with individual coronavirus replicase proteins, and a number of cellular proteins have also been implicated as components of the MHV replicase-transcriptase complex [34–36]. However, the essential nature of some of these cellular proteins has been questioned [37], and further work is needed to determine the exact protein composition of the coronavirus replicase-transcriptase complex and how the composition is altered, or how the proteins are modified to regulate the different activities of the complex.
In order to address these sorts of questions, we have embarked upon a detailed analysis of temperature-sensitive (ts) mutants of MHV-A59 that are unable to synthesize viral RNA when the infection is initiated and maintained at the non-permissive temperature. The essential feature of these mutants is that they are likely to be defective in different aspects of viral RNA synthesis and a detailed characterization of their genotype and phenotype should provide insights into the mechanisms of RNA synthesis, the functions of individual viral replicase proteins, and the protein-RNA and protein–protein interactions that regulate the activity of the replicase–transcriptase complex. These conditional-lethal mutants may also be used in a cis–trans test to define the number of complementation groups, or cistrons, that contribute to a specific phenotype. This sort of analysis can also provide valuable insight into the possible pathways that polyproteins must travel to assume functional configurations and has been used with success for other RNA viruses [38].
The MHV-A59 mutants that we study have been produced in a number of laboratories over a period of 20 years [39–41]. They have been selected to have a low efficiency of plaque formation at the non-permissive temperature compared with the permissive temperature and hence a reversion frequency indicative of single point mutations. In this study, we describe a complementation analysis, and by sequence analysis of both ts virus and revertants, we identify the causal mutation for eight of these mutants. We also describe a more detailed phenotype for selected mutants and suggest a model that describes the different modes of RNA synthesis during coronavirus replication and transcription.
Results
Characterization of ts Mutants and Revertants
Table S1 lists the ts mutants of MHV-A59 used in our collection. All the ts mutants failed to form plaques or synthesize viral RNA when infection was initiated and maintained at the non-permissive temperature. While many mutants failed to form plaques at 37 °C, other mutants formed plaques at 37 °C and were considered leaky. This included Alb ts22 that produced pin-prick-sized plaques after 2 d at 37 °C (compared with the wild-type [wt] A59 virus, which produced uniform plaques of 4–5 mm in diameter) and Wü ts18, Wü ts36, and Wü ts38, which produced smaller than wt plaques at 37 °C. However, even for these mutants, the ts defects responsible for their RNA-negative phenotype appeared to be caused by a single point mutation because each ts mutant possessed a characteristic low reversion frequency between 10−4 and 10−8 per average base [42]. The virus produced at 37 °C by Alb ts22, Wü ts18, Wü ts36, and Wü ts38 was also ts, i.e., the efficiency of plating (EOP) was less than 10−4.
For most mutants, the revertant virus obtained from plaques formed at the non-permissive temperature had properties identical to wt MHV-A59. One exception was Alb ts17, which produced equal numbers of revertant viruses causing A59-sized plaques and revertant viruses with noticeably smaller plaques (Figure S1). We isolated revertant viruses from a large (A59-sized) plaque (Alb 17RL) and a small plaque (Alb 17RS) for sequence analysis (see below). Some of the ts mutants did not produce revertant viruses (e.g., LA ts3, Alb ts19) or produced revertant viruses that were markedly different from the parental MHV-A59 virus.
Complementation Analysis
We began our complementation analyses using Alb ts16, LA ts6, and Alb ts22 because they each had a distinct ts viral RNA synthesis phenotype (see below). Cells were singly infected or doubly infected with two ts mutants and the cells and medium were harvested after the completion of a single round of replication, i.e., 8 h post-infection (hpi) at 40 °C. We also confirmed that if infection with a ts virus alone was allowed to proceed for up to 2 h at 30 °C, and then the culture shifted to 40 °C and the virus harvested at 12 hpi, the titer we obtained was low (~104 plaque-forming units [pfu]/ml). Thus, this protocol prevented the production of revertant virus by a second round of replication. Complementation was measured by determining the complementation index (CI) as described in Materials and Methods. By definition, if the mutations are in the same cistron, the viruses will not complement each other. On the other hand, if the mutations are in different cistrons, the mutants will complement each other and progeny ts virus will be recovered.
The results of six individual crosses between Alb ts16 and LA ts6 are shown in Table 1. All of these crosses failed to show complementation. The average CI value was 0.5 (0.5 ± 0.18 SD), which is the theoretical value for two mutants with mutations in the same cistron [43]. This CI value was obtained using only the titers determined at 30 °C and was not corrected for the presence of revertants (or recombinants) as was done by others [39,44]. We found it unnecessary to make this correction because it did not significantly change the CI value (at most a decrease of one tenth) and whether or not the mutants scored as able to complement one another. From these results, we concluded that Alb ts16 and LA ts6 had a mutation in the same cistron and were, therefore, in the same complementation group. We next determined if Alb ts22 would complement Alb ts16 or LA ts6. As shown in Table 1, in three separate experiments Alb ts22 clearly complemented both Alb ts16 and LA ts6. Therefore, the mutation in Alb ts22 was in a different cistron than the mutations in Alb ts16 and LA ts6, thus identifying a second complementation group. In a series of further experiments, we extended our complementation analysis to include Alb ts6, Wü ts18, Wü ts36, and Wü ts38. Using the same assay, we found that Alb ts6 complemented Alb ts22 but failed to complement Alb ts16 or LA ts6. Thus, we conclude that Alb ts6 was in the same complementation group as Alb ts16 and LA ts6. Finally, we found that Wü ts18, Wü ts36, and Wü ts38 were in a different complementation group(s) from that of either Alb ts6 or Alb ts22, and thus, these mutants defined at least a third complementation group.
Table 1 Genetic Complementation Analysis of MHV-A59 ts Mutants
In our analysis of the ts mutants of MHV-A59 described above, values for the CI were always less than two or more than five and thus readily interpreted as positive or negative without correction for the presence of revertants or recombinants. However, from the results we obtained, it was clear that recombination did occur when there was complementation. The EOP of the virus harvested from cells co-infected with two complementing viruses was usually ~10−2, and not the EOP of the individual ts mutants, which was 10−4–10−8. This result is in contrast to similar experiments using Sindbis virus in complementation assays, where we obtained similar EOPs to the input viruses when assaying the progeny from complementing ts mutants (unpublished data). We took these results to indicate that complementation allowed recombination in MHV.
This finding provided the means to develop a more convenient and more rapid method of determining complementation for MHV-A59 ts mutants. We reasoned that because recombination appeared to be driven by complementation, biochemical complementation (i.e., viral RNA synthesis) might be detected in cells co-infected with complementing ts mutants, but not in cells infected with ts mutants in the same complementation group. We devised such an assay. Cells were infected at the permissive temperature and were then re-fed with medium prewarmed to the non-permissive temperature and containing dactinomycin to inhibit DNA-dependent RNA synthesis and 3H-uridine to label viral RNA. The infected cells were incubated until 7–8 hpi at 39 °C to 40 °C or 8–12 hpi at 30 °C, and RNA synthesis was measured by the incorporation of 3H-uridine into acid-precipitable material. Figure 2A shows the results of single and double infection with the Alb ts6, Alb ts16, Alb ts22, and LA ts6 mutants. The data show that at 40 °C, the mutants Alb ts6, Alb ts16, and LA ts6 were not able to rescue the RNA-negative phenotype of each other and thus, the three mutants were in the same complementation group. In contrast, Alb ts22 was able to rescue the RNA-negative phenotype of Alb ts6, Alb ts16, and LA ts6 and thus, was the sole member of a separate complementation group. This result is identical to that obtained using classical complementation assays and served to validate the new method. The assay was as specific as classic genetic complementation, which measures progeny virus production, but is less time-consuming.
Figure 2 Biochemical Complementation Analysis of Selected MHV-A59 ts Mutants
Cells were mock-infected or infected with MHV-A59, one of the ts mutants, or with a mixture of two ts mutants. The cells were incubated at 40 °C in medium containing dactinomycin and 3H-uridine and, at 8 hpi, 3H-uridine incorporation into trichloroacetic acid-precipitated RNA was determined. Cells were infected with: M, mock-infected; A59, MHV-A59; A6, Alb ts6; A16, Alb ts16; A22, Alb ts22; A17, Alb ts17; L6, LA ts6; W18, Wü ts18; W36, Wü ts36; W38, Wü ts38; A6xA16, Alb ts6 and Alb ts16; A6xL6, Alb ts6 and LA ts6; A6xA22, Alb ts6 and Alb ts22; A16xL6, Alb ts16 and LA ts6; A16xA22, Alb ts16 and Alb ts22; L6xA22, LA ts6 and Alb ts22; A17x A16, Alb ts17 and Alb ts16; A17xL6, Alb ts17 and LA ts6; A17xA22 or A22xA17, Alb ts17 and Alb ts22; A17xW38, Alb ts17 and Wü ts38; A17xW18, Alb ts17 and Wü ts18; A17xW36, Alb ts17 and Wü ts36; A22xW18, Alb ts22 and Wü ts18; A22xW36, Alb ts22 and Wü ts36; A22xW38, Alb ts22 and Wü ts38; W18xW36, Wü ts18 and Wü ts36; W18xW38, Wü ts18 and Wü ts38; W36xW38, Wü ts36 and Wü ts38.
Using this assay, we were able not only to confirm the prediction of at least three complementation groups that were obtained using classical complementation procedures but also to identify a fourth complementation group. The results are presented in Figure 2B and 2C and show that mutants Alb ts17, Wu ts36, Wu ts38, and Wu ts18 define not one but two additional complementation groups. We found Alb ts17 and Wü ts38 belong to the same complementation group based on their failure to complement each other's defects. However, both of these mutants complemented Wü ts36 and Wü ts18, which did not complement each other.
Finally, we extended this assay to include the full collection of mutants that we have available and Table 2 summarizes the complementation patterns of the RNA-negative ts mutants of MHV-A59 assayed to date. The numbers shown in Table 2 represent the percentage of viral RNA synthesis found for the mixed mutant-infected cells compared to A59 virus-infected cells at 40 °C. A value less than zero means the 3H-uridine incorporation was less than that obtained from mock-infected cells. With this type of assay, we took less than 1% of the MHV-A59 incorporation as indicating failure to complement and greater than 1% as evidence of positive complementation. Based on these results, it was possible to assign a further ten mutants (Alb ts2, ts8, and ts9, and ts19; Ut ts88 and ts329; LA ts3 and ts9; and NC ts2 and ts3) to the same complementation group as Alb ts6, Alb ts16, and LA ts6 and, it was possible to assign mutant Ut ts145 to the same complementation group as Wü ts18 and Wü ts36. Thus, it was possible to assign the entire collection of 19 RNA-negative ts mutants of MHV to one of four complementation groups, which we have tentatively named cistrons I, II, IV, and VI based on the locations identified for their causal point mutations (see below). This numbering scheme leaves open the possibility of finding two additional complementation groups (cistrons III and V) in the future that would represent gene products of ORF1b (see below).
Table 2 Biochemical Complementation Analysis of MHV-A59 ts Mutants
Identification of Mutations Responsible for the ts Mutant Phenotype
The entire coding region of the replicase genes (ORF1a and ORF1b) was sequenced for each of eight ts mutant/revertant pairs. In each case, a single nucleotide change was identified as the mutation responsible for the ts mutant phenotype. Using the numbering that we have assigned to the infectious cDNA clone of the MHV-A59 genome [45] (GenBank accession number AY700211), the nucleotide changes compared to wt MHV-A59 were identified as: Alb ts6, A9494→ C; Alb ts16, U10864→ C; LA ts6, C13360→ G; Alb ts22, A16180→ G; Alb ts17, G19288→ A; Wü ts38, U19383→C; Wü ts18, C20880→U; Wü ts36, U21304→ C (Figure 3A). We also identified a number of nucleotide differences between mutants isolated in different laboratories, but in no case did they correlate with the ts phenotype. With the exception of the Alb 17RS revertant, all of the revertants we isolated were true, i.e., they were genetically and phenotypically identical to the wt MHV-A59. The Alb 17RS revertant was a pseudorevertant in that the nucleotide at position 19288 had reverted from A→ C, which resulted in a substitution of Tyr with Arg. This radical substitution was reflected in a small plaque phenotype (Figure S1).
All of the nucleotide changes responsible for the ts mutant phenotype were non-synonymous mutations. The amino acid substitutions are shown in Figure 3B. Conservative substitutions were identified in nsp5 and nsp10 of the Alb ts16 and LA ts6 mutants, respectively. Moderately conservative substitutions were identified in nsp4 and nsp12 of the Alb ts6 and Alb ts22 mutants, respectively. And radical substitutions were identified in nsp14 of the Alb ts17 and Wü ts38 mutant, as well as nsp16 of the Wü ts18 and Wü ts36 mutants [46]. A comparison of the predicted replicase protein sequences from different coronaviruses showed that there was, by and large, conservation of the amino acids that were substituted in the proteins with a ts phenotype. For example, the Gln65 residue of nsp10, the His868 residue of nsp12, and the Cys408 residue of nsp14 appear to be well conserved in Group I, II (including SARSCoV), and III coronaviruses. In contrast, the Asn258 residue of nsp4 is only found in MHV strains, although in the majority of other coronaviruses, it is substituted by an aspartic acid. Finally, it is possible, with different degrees of confidence, to predict the structural environment in which the residues in question are found. On the one hand, it is highly likely that the Phe219 residue of nsp5 is located in an extended area that connects the α-helices B and C in the carboxyl-terminal domain III of nsp5, the 3C-like proteinase. This conclusion is based upon the similarity in the sequences of coronavirus nsp5 proteins and the crystallographic structures that have been solved for the transmissible gastroenteritis virus (TGEV), SARSCoV, and HCoV-229E nsp5 proteins [23,47,48]. On the other hand, programs that predict protein secondary structure [49] indicate that the Gln65 residue of nsp10, the His868 residue of nsp12, and the Cys408 residue of nsp14 are located in disordered loop structures, while the Asn258 residue of nsp4 and the Leu153 residue of nsp16 are involved in α-helices. Obviously, more definitive structural data will be needed to confirm these predictions.
Figure 3 Genotypic Analysis of Selected MHV-A59 ts Mutants
(A) The positions of mutations responsible for the ts phenotype of selected MHV-A59 mutants are illustrated in relation to the non-structural proteins (nsp1–16) produced by proteolytic processing of the ORF1a/ORF1b polyprotein, pp1ab. Nucleotide changes are numbered according to the sequence of the infectious cDNA clone of MHV-A59.
(B) The amino acid substitutions responsible for the mutant and revertant phenotypes are listed together with the mutated protein and the cistron to which each mutant has been assigned. The amino acids are numbered from the amino-terminus to the carboxyl-terminus of each of the non-structural proteins.
Phenotypes of the MHV-A59 ts Mutants
We focused our phenotypic analysis on the eight MHV-A59 ts mutants that had been genotyped and began by measuring “total” viral RNA synthesis in infected cells prior to and following shift from the permissive to the non-permissive temperature. This analysis was done after 8 h of incubation at 30 °C, a time at which the replicase-transciptase complex produces mainly (>90%) positive-strand RNA, and ~20% of the maximum rate of RNA synthesis has been reached. Mutant virus-infected cells were shifted to 40 °C at 8 hpi and a duplicate set was left at 30 °C. Both sets of cultures were labeled for 1 h with 3H-uridine in the presence of 10 μg per ml of cycloheximide (CH) to monitor the replicase-transcriptase activity at the time of shift. The results are shown in Figure 4. In MHV-A59 infected cells, the amount of 3H-uridine incorporation doubled, as expected, when the temperature was increased by 10 °C. The group I mutants had about the same level of viral RNA synthesis at both temperatures, while in the group II, IV, and VI mutant-infected cells, viral RNA synthesis diminished by 50% or more in the hour following temperature shift. We interpret this to mean that mutations in replicase proteins encoded in ORF1a appeared to confer temperature-sensitivity to the viral replicase-transcriptase complex, but once it had formed at 30 °C, its positive-strand synthetic activity was relatively resistant to higher temperature. In contrast, mutations in ORF1b-encoded proteins, namely nsp12, nsp14, and nsp16 appeared to affect the positive-strand synthetic activity of already-formed replicase-transcriptase complexes. We then went on to analyze the phenotypes of three ts mutants in more detail.
Figure 4 RNA Synthesis Phenotype of MHV-A59 ts Mutants
RNA synthesis was determined using a 1 h pulse label with 3H-uridine in the presence of dactinomycin and cycloheximide, given to wt MHV-A59 and ts mutant virus-infected cells at 8 hpi with or without shifting from the permissive to the non-permissive temperature. The amount of incorporated 3H-uridine at 40 °C was divided by that at 30 °C and 1.0 was subtracted. The results represent the average of five separate experiments. A value of zero means the incorporation at the two temperatures was the same.
Alb ts22.
The phenotype described above for group II, IV, and VI mutants would be consistent with a defect in any stage of positive-strand RNA synthesis. In the case of mutant Alb ts22, however, we have shown that the ts lesion is located in nsp12, the viral RNA-dependent RNA polymerase subunit. This suggested to us that the Alb ts22 might be defective in the elongation phase of RNA synthesis. To analyze the phenotype of Alb ts22 in more detail, RNA synthesis in Alb ts22-infected cells was determined using 1 h pulse labels with 3H-uridine in the presence of dactinomycin, given between 1–6 hpi at 40 °C or between 5–14 hpi at 30 °C (Figure 5A). At 40 °C, Alb ts22-infected cells incorporated only mock levels of 3H-uridine, as expected for an RNA-negative ts mutant. In contrast, cells infected with wt MHV-A59 or with Alb 22R (a revertant of Alb ts22) made RNA at high rates and at identical times. At 30 °C, Alb ts22 was defective in viral RNA synthesis and never reached the levels of viral RNA synthesis shown by wt MHV-A59 or Alb 22R. These results are consistent with our finding that, at 30 °C, the plaques formed by Alb ts22 were smaller that those formed by wt MHV-A59. Analysis by gel electrophoresis of the species of positive-strand RNA made in Alb ts22-infected cells at 30 °C showed the typical pattern of seven RNAs (genome and six subgenomic mRNAs), although the six subgenomic mRNAs were reduced equally in amount relative to the genome RNA when compared to Alb 22R infected cells (unpublished data). We conclude that Alb ts22 not only produced less overall RNA compared to wt MHV-A59 and Alb 22R, even at the permissive temperature, but also under-produced all of the subgenomic mRNA species relative to the genome RNA.
Figure 5 RNA Synthesis Phenotype of the Alb ts22 Mutant
RNA synthesis was determined (A) using 1 h pulse labels with 3H-uridine in the presence of dactinomycin, given to MHV-A59-, Alb ts22-, and Alb 22R-infected cells 1–6 hpi at 40 °C or 5–14 hpi at 30 °C; █, 40 °C, wt MHV-A59; ▴, 40 °C, Alb 22R; •, 40 °C, Alb ts22; □, 30 °C, wt MHV-A59; ▵, 30 °C, Alb 22R; ○, 30 °C, Alb ts22, or (B) using 30 min pulse labels with 3H-uridine in the presence of dactinomycin, given to MHV-A59-, Alb ts22-, and Alb 22R-infected cells after shift from the permissive to the non-permissive temperature at 13 hpi; █, wt MHV-A59; ▴, Alb 22R; •, Alb ts22.
We also examined the ability of Alb ts22-infected cells to continue viral RNA synthesis after shift from 30 °C to 40 °C at 13 hpi (Figure 5B). This allowed us to follow the activity at 40 °C of the viral RNA-dependent RNA polymerase that was made and assembled at 30 °C. At this time, Alb ts22 RNA synthesis was at its maximum rate and RNA synthesis by wt MHV-A59 and Alb 22R was declining. The results show that a shift to 40 °C led to the rapid loss of RNA synthesis by Alb ts22 but not by wt MHV-A59 or Alb 22R. This result is consistent with a failure of the viral RNA-dependent RNA polymerase to continue transcription at the non-permissive temperature. We concluded Alb ts22 had a ts defect in elongation, although we do not know if elongation is directly affected or if the amino acid change in nsp12 affects its interaction with an as yet unknown, but essential protein. We have also shown that, as expected, Alb ts22 is unable to synthesize negative-strand RNA at the non-permissive temperature (unpublished data).
Alb ts16 and LA ts6.
Although both Alb ts16 and LA ts6 are unable to synthesize viral RNA when the infection is initiated and maintained at the non-permissive temperature, the data shown in Figure 4 suggests that they are not significantly impaired in their ability to synthesize positive-strand RNA at this temperature. This conclusion is strengthened by the results shown in Figure 6A, which demonstrate the kinetics of overall viral RNA synthesis in Alb ts16 and LA ts6 virus-infected cells after shifting the incubation temperature from 30 °C to 40 °C at 8 hpi. With wt MHV-A59, viral RNA synthesis increased rapidly within the first 60 min after temperature shift, consistent with the synthesis of both additional negative-strand templates and their nascent positive-strand product. The addition of CH at the time of shift resulted in a constant rate of viral RNA synthesis for at least 1 h. As we know that negative-strand synthesis in MHV-A59-infected cells is short-lived and stops within 30 min of the inhibition of protein synthesis [24], we deduce that the addition of CH prevented the synthesis of new viral proteins, which in turn prevented the formation of additional replicase-transcriptase activity and negative-strand templates.
Figure 6 RNA Synthesis Phenotype of the Alb ts16 and LA ts6 Mutants
RNA synthesis (A) or negative-strand RNA synthesis (B) was determined using 20 or 30 min pulse labels with 3H-uridine in the presence of dactinomycin, with or without the addition of CH, after shifting the incubation temperature of MHV-A59-, Alb ts16-, and LA ts6-infected cells from 30 °C to 40 °C at 8 hpi: filled bar, 0–20 min pulse; grey bar, 20–40 min pulse; open bar, 40–60 min pulse; dark diagonal bar, 0–30 min pulse; light diagonal bar, 30–60 min pulse.
In cells infected with complementation group I ts mutants Alb ts16 and LA ts6, viral RNA synthesis continued at 40 °C at the level ongoing at the time of temperature shift (Figure 6A). This meant that the replicase-transcriptase complexes assembled at 30 °C continued to function at 40 °C in the synthesis of positive-strand RNA. However, unlike A59-infected cells, the group I mutants did not increase their rates of RNA synthesis after shift to non-permissive temperature, indicating that no new active complexes were formed. This phenotype resembled that seen with MHV-A59-infected cells treated with CH, and we conclude that the complementation group I mutants are defective in their ability to form active replicase-transcriptase complexes at 40 °C but retain the positive-strand synthesis activity of the complexes formed at the permissive temperature.
At least two possibilities could account for a failure of group I ts mutants to form fully competent replicase-transcriptase complexes at the non-permissive temperature. Either no new negative-strand templates were made, i.e., a defect in negative-strand synthesis, or, if they were made, they could not be used as templates for positive-strand synthesis. The latter phenotype has been observed for certain alphavirus mutants [50], which were called conversion-defective mutants. To distinguish between these two possibilities, it is necessary to shift the ts mutant-infected cells to the non-permissive temperature and determine their ability to continue negative-strand RNA synthesis. Mutants that fail to continue negative-strand RNA synthesis would be defective in this step, while mutants that continued to make negative strands would be designated as conversion-defective mutants. Cells infected with wt MHV-A59, Alb ts16, and LA ts6 were shifted from 30 °C to 40 °C at 8 h (Alb ts16) or 9 h (LA ts6) post-infection and were pulse-labeled with 3H uridine at 40 °C. Then, viral negative-strand templates in replicating and transcribing structures were purified free of single-stranded RNA, and the incorporation of radioactivity into negative-stranded RNA was measured by nuclease protection assays. In this assay, the results are expressed as the percentage of the 3H-uridine incorporated into the negative-stranded component of the purified, nuclease-resistant RNA cores of the replicative-transcriptive structures. As these structures represent double-stranded RNA, if 40%–50% of the total incorporation in the core RNA is found in negative strands, it means that 80%–100% of the negative strands that were active as templates during the pulse period had been made during this same period. This occurred when negative-strand synthesis was measured early in the infection cycle, when viral RNA synthesis was ~20% of the maximum [24].
Figure 6B shows that in wt MHV-A59-infected cells, negative-strand synthesis continues following a shift from 30 °C to 40 °C at a time when the amount of viral RNA synthesis is ~20% of maximum. This is seen by the similar high values of 20%–25% of the labeled RF RNA being in negative strands for successive 30 min pulse-periods in the absence of CH. Also, as shown previously, continued negative-strand synthesis in MHV-A59-infected cells is dependent on continued translation and abruptly declines in the presence of CH. In the case of LA ts6, the percentage of 3H-uridine incorporated in negative strands declined abruptly after shifting to 40 °C and this decline was the same in the absence or the presence of CH. With Alb ts16, negative-strand synthesis continued during the 20–40 min and the 40–60 min pulse-periods in the absence of CH but was inhibited in the presence of CH. For this mutant, to find that negative-strand synthesis continued at 40 °C without an increase in the rate of positive- strand synthesis, as seen for MHV-A59, was consistent with Alb ts16 having a ts defect affecting the ability of the negative-strand templates to be efficiently used at 40 °C. Thus, we conclude that LA ts6 was defective in continuing negative-strand synthesis after shift to 40 °C and Alb ts16 displayed what appears to be a conversion phenotype.
Discussion
Taken together with the complementation analysis, the identification of the mutations responsible for the ts phenotypes of Alb ts6, Alb ts16, Alb ts17, Alb ts22, LA ts6, Wü ts18, Wü ts36, and Wü ts38 leads to a number of important conclusions. First, our data strongly suggest that most of the replicase gene products of ORF1a are cis-active and form a single complementation group (cistron I) encompassing, at least, the nsp4 to nsp10 coding region. If correct, our conclusion must mean that a large proportion of nsp1–nsp11 proteins function as a polyprotein, if only initially or transiently, or they associate as a cis-acting complex before they are proteolytically processed. We favor a model in which a pp1a-related polyprotein represents a large modular scaffolding protein that displays binding sites for ORF1b-encoded pp1ab processing products. While the large number of mutants that fall into cistron I clearly suggest that it is extensive and polygenic, it is not yet clear if all of the ORF1a-encoded proteins function in cis. We are aware that the arterivirus Equine arteritis virus ORF1a-encoded protein nsp1 can function in trans [51] and it has recently been shown that the MHV-A59 ORF1a-encoded protein nsp2 is non-essential for virus replication [52]. The genetic analysis of further MHV-A59 ts mutants will be needed to define the precise boundaries of MHV-A59 cistron I.
Second, our results suggest that the replicase gene products encoded in ORF1b (i.e., nsp12–nsp16) are diffusible and thus assemble and function in viral RNA synthesis after cleavage from pp1ab. This also leads us to the prediction that there will be five cistrons in ORF1b, each corresponding to one of the proteolytic cleavage products, and we have designated them tentatively as cistrons II–VI in a 5′ to 3′ direction (nsp12, cistron II; nsp13, cistron III; nsp14, cistron IV; nsp15, cistron V; and nsp16, cistron VI). The idea that the MHV-A59 ORF1b-encoded replicase proteins function in trans is consistent with the results of Brockway et al., who have shown that a green fluorescent protein–tagged MHV-A59 nsp12 is able to diffuse into the replicase-transcriptase complex if expressed individually in virus-infected cells [9]. However, we would also like to note that our data does not exclude the possibility that some of the ORF1b-encoded proteins may function as intermediates, rather than the end products of proteolytic cleavage. For example, functional proteins comprising nsp12/nsp13, nsp13/nsp14, nsp14/nsp15, nsp15/nsp16 as well as nsp13/nsp14/nsp15 could all be accommodated as single cistrons based upon our complementation data. This would lead to the prediction of either three or four cistrons encoded in ORF1b. The idea that a number of the enzymes involved in coronavirus RNA synthesis may be linked not only functionally, i.e., sequentially in a concerted reaction pathway, but also structurally (i.e., as multifunctional proteins) is also suggested by other studies. For example, Ziebuhr and colleagues [53] have shown that 2′-O-ribose-methylated RNA substrates are resistant to cleavage by the SARS-coronavirus endoribonuclease (nsp15), indicating a functional link with the S-adenosylmethionine-dependent 2′-O-methyl transferase (nsp16). We are currently searching for further ts mutants that might help resolve this issue and we are attempting to trans-complement ts mutants with cell lines that constitutively express ORF1b-encoded replicase proteins. Despite these reservations, the genetic data do allow us to conclude that not only nsp5, the 3C-like cysteine proteinase, and nsp12, the putative RNA-dependent polymerase (as might have been predicted), but also nsp14, the putative MHV exonuclease, nsp16, the putative MHV 2′-O-methyltransferase, nsp4, and nsp10 are essential for the assembly of a functional replicase-transcriptase complex.
In contrast to most other positive-stranded RNA virus, the viral replicase-transcriptase complex of coronaviruses (and most other nidoviruses) functions to amplify the genome via a full-length negative-strand intermediate and to produce, via a discontinuous process, subgenome-length negative-strand templates that are then copied directly into subgenomic mRNA. How the replicase-transcriptase complex accomplishes these various activities is not understood in any detail. For example, it is not known whether the same replicase-transcriptase complex functions to produce full-length and subgenome-length RNA or how the conversion from negative- to positive-strand RNA synthesis is regulated. Does the analysis of MHV-A59 ts mutants help us to understand these complex processes?
We have shown previously that negative- and positive-strand RNA synthesis occurs simultaneously throughout MHV-A59 infection but that negative-strand synthesis is short-lived, i.e., its synthesis halts within several minutes after protein synthesis is inhibited [24]. This contrasts with positive-strand synthesis, which continues unabated for 1 h and then gradually declines and disappears about 4 h after the inhibition of translation. These observations suggest that unprocessed forms of the replicase polyprotein(s) might function in negative-strand synthesis and that cleavage of the nascent polyprotein inactivates the negative-strand activity of the replicase, as it does for alphaviruses [54,55]. The replicase-transcriptase activity for positive-strand synthesis can be restarted after the block of translation is reversed [27] but, for this to happen, new negative-strand templates need to be synthesized. In other words, it appears that the coronavirus replicase-transcriptase complex ages, losing both its negative-strand templates and its activity. This interpretation fits well with our genetic analysis of the mutants LA ts6, Alb ts16, and Alb ts6, which shows that they all fall into a single complementation group. It is also consistent with our proposal that the replicase proteins encoded in ORF1a are expressed and function as a polyprotein, or that they assemble as a cis-acting complex before they are proteolytically processed. It is also worth noting that in vivo protein labeling experiments indicate that proteolytic processing of both MHV-A59 ORF1a and MHV-A59 ORF1b-encoded replicase proteins is measured in hours rather than minutes [56–58] and that the fully processed 3C-like cysteine proteinase is first detected several hours post-infection [59], a time at which the rate of viral RNA synthesis is already increasing rapidly [24].
The idea that the MHV replicase-transcriptase complex is active in negative-strand RNA synthesis before pp1a is extensively processed also fits well with our detailed phenotypic analysis of cistron I mutants. In the case of LA ts6, negative-strand synthesis was inhibited after shift to the non-permissive temperature and, in time, this leads to a decline in positive-strand RNA synthesis (unpublished data). Thus, at the non-permissive temperature, LA ts6 could not sustain positive-strand synthesis, nor replace or replenish aging replicase-transcriptase complexes. The causal mutation in LA ts6 would substitute a Glu for the Gln65 residue of wt nsp10. As noted above the Gln65 residue is conserved in Group I, II (including SARSCoV), and III coronaviruses and its substitution with Glu might prevent the proper folding of pp1a into a conformation that would allow it to participate in the formation of a replicase-transcriptase complex with negative-strand activity. It would be interesting to determine if, at the non-permissive temperature, nsp10 of LA ts6 could associate with nsp12, nsp13, nsp14, nsp15, or nsp16. Also, it was curious that LA ts6 had a very low reversion frequency of ~10−8. Why certain bases fail to revert at the typical frequency of 10−4 to 10−5 is unknown but may be indicative of a region of the genome that is transcribed with higher fidelity than other regions. Alternatively, this low reversion frequency may be an inherent property of the LA ts6 replicase-transcriptase complex.
In contrast to LA ts6, Alb ts16 appeared to be able to continue to form negative strands after shift to the non-permissive temperature, but these negative strands were not converted into templates for positive-strand synthesis. We speculate that Alb ts16 has a ts defect in the conversion of the replicase-transcriptase complex from one able to synthesize negative strands to one able to synthesize positive strands. It is certainly suggestive that Alb ts16 had a mutation in nsp5, which is the 3C-like proteinase of the virus, but it has yet to be determined if this mutation affects the activity of the proteinase, or if it affects the folding of pp1a or pp1ab, or if the nsp5 C-terminal domain itself could have a function in positive-strand RNA synthesis. Nevertheless, because negative-strand RNA synthesis was inhibited in Alb ts16-infected cells treated with CH at the time of shift to non-permissive temperature, we propose that the Alb ts16 replicase-transcriptase complex does not retain its activity for minus-strand synthesis. Rather it fails to gain positive-strand synthesis activity at the non-permissive temperature. We favor a model where the activity that makes positive strands is gained at the expense or loss of the activity to make negative strands.
Finally, although we are able to rationalize the genotype of Alb ts22, i.e., a mutation in nsp12 (the RNA dependent RNA polymerase) with its phenotype (i.e., an immediate effect on RNA synthesis at the non-permissive temperature) we were surprised to find that Alb ts17, Wü ts18, Wü ts36, and Wü ts38 also showed the same phenotype but had mutations in other replicase proteins. Generally, it is unusual to find so many ts mutants that show an effect on RNA synthesis if the replicase-transcriptase complex is first allowed to assemble at the permissive temperature. Most RNA-negative ts mutants of alphaviruses, for example, fail to make viral RNA when the infection is initiated at the non-permissive temperature but continue to make viral RNA if shifted to the non-permissive temperature late in infection (unpublished data). One possibility is that nsp14 and nsp16 dissociate or become less tightly associated with the replicase-transcriptase complex after shifting to the non-permissive temperature and this causes the complex to lose elongation activity. Another possibility is that the enzymatic activities associated with nsp14 and nsp16 are altered in the group IV and group VI mutants. Further studies will be required to explain this phenotype.
In summary, our detailed phenotypic analysis of MHV-A59 ts mutants allows us to propose a working model that describes a pathway for viral RNA synthesis in MHV-A59-infected cells. In this model, the replicase-transcriptase complex forms initially and creates a negative-strand template. It is then converted to utilize the negative strand as a template for positive-strand synthesis and, finally, the complex is inactivated by the degradation of negative-strand templates (Figure 7). Our analysis also allows us to place some of our ts mutants at specific points on this pathway. We hope that a more detailed biochemical analysis of these mutants will allow us to identify intermediates in the pathway of RNA synthesis and will provide valuable information of the precise function(s) of the viral replicase proteins involved. Furthermore, we believe that the characterization of these mutants provides an excellent starting point for the generation of second site reversion mutants. This could be done, for example, by using the recently developed MHV reverse genetic system [45] to generate ts mutants with codon, rather than nucleotide substitutions. Second site reversion mutants may then provide valuable information on protein-protein interactions within the replicase-transcriptase complex.
Figure 7 A Model to Describe the Pathway for Viral RNA Synthesis in MHV-A59-Infected Cells
Shows a working model that describes a pathway for viral RNA synthesis in MHV-A59-infected cells. The model proposes that the replicase-transcriptase complex forms initially and creates a negative-strand template. It is then converted to utilize the negative strand as a template for positive-strand synthesis and, finally, the complex is inactivated by the degradation of negative-strand templates. It is also proposed that proteolytic processing of the replicase polyproteins plays a role in regulation of this pathway. Also depicted are the putative defects of specific MHV-A59 ts mutants. It remains to be shown whether or not the group IV and VI mutants (Wü ts38, Alb ts17, Wü ts18, and Wü ts36) are defective in negative-strand RNA synthesis at the non-permissive temperature.
Materials and Methods
Cells and viruses.
Seventeen clone one (17Cl-1) mouse fibroblast cells [60] were cultured at 37 °C in Dulbecco's modified Eagle's medium (DMEM) supplemented with 6% fetal bovine serum (FBS), 5% tryptose phosphate broth (TPB), penicillin (100 units/ml), and streptomycin (100 μg/ml). Sac(−) cells [61] were cultured at 37 °C in minimal essential medium (MEM) supplemented with 5% FBS, penicillin (100 units/ml), and streptomycin (100 μg/ml). The A59 strain of MHV and a set of ts mutants derived from MHV-A59 (Alb prefix) were originally obtained from the laboratory of L. Sturman, Wadsworth Center for Laboratories and Research, Albany, New York, United States [40]. Mutants prefixed with LA (Los Angeles) and NC (North Carolina) were obtained from M. Schaad and R. Baric, University of North Carolina, Chapel Hill, North Carolina, United States and have been initially characterized [39]. Mutants prefixed with Ut (Utrecht) were obtained from W. Spaan, Leiden University Medical Center, Leiden, The Netherlands and have been initially characterized [41]. The LA, NC, and Ut ts mutants were derived from different but related lineages of the Albany isolate of MHV-A59. Mutants prefixed with Wü (Würzburg, Germany) were isolated as described below.
For our studies, virus stocks were derived from the original mutant isolates after plaque purification and propagation in 17Cl-1 cells cultured at 30 °C or 33 °C in low pH DMEM (pH 6.4) containing 6% FBS, 5% TPB, penicillin (100 units/ml), and streptomycin (100 μg/ml) [24]. Revertants were picked from plaques of mutants titered at 39.5 °C, followed by another plaque purification at 39.5 °C. The virus stocks used were first passage and were obtained by using virus from a single plaque (~107 pfu) to infect a dish of ~4 × 107 17Cl-1 cells to yield 30 ml of stock virus with a titer of 1–6 × 109 pfu/ml.
Isolation of Wü ts mutants.
Sac(−) cells were infected with 10 pfu/cell of MHV-A59 (originally obtained from P. Carthew, Medical Research Council Laboratories, Carshalton, United Kingdom), and incubated for 15 h at 37 °C in medium containing 150 μg/ml of 5-fluorouracil. This concentration of pyrimidine analogue was determined to inhibit virus replication by 80%. The mutagenized virus stock was diluted to 15 pfu/ml in medium and 100 μl aliquots were incubated with 104 Sac(−) cells at 30 °C for 48 h. The supernatant was taken from cultures that displayed syncytium formation and used to infect duplicate cultures of 104 Sac(−) cells that were incubated at 30 °C or 39.5 °C for 24 h. The supernatant was taken from replica cultures that developed cytopathic effect at 30 °C but not at 39.5 °C, and potential ts mutants were isolated by two plaque purifications. Sequence analysis of the Würzburg strain of MHV-A59 suggests that approximately 8,000 nucleotides at the 5′ end of the genome have been exchanged by recombination with a related but different MHV strain (unpublished data).
Characterization of mutant stocks for titer and EOP.
17Cl-1 cells in 60 mm petri dishes were infected with 0.5 ml of 10-fold dilutions of virus at room temperature. Virus was diluted in an infection medium (MEM with Hank's balanced salts [HBSS] containing 50 μg/ml of DEAE-dextran [diethylaminoethyl dextran], 0.2% bovine serum albumin, and 20 mM HEPES [N-2-hydroxyethylpiperazine-N′-2-ethansulfonic acid] [pH 6.6]). The inoculum was removed after 30 min and the cells were overlaid with DMEM containing 6% FBS, penicillin (100 units/ml), streptomycin (100 μg/ml), and 0.1% Gelrite™ gellan gum (Sigma, St. Louis, Missouri, United States) and incubated at the appropriate temperature in 7.5% CO2. After incubation for 3 or 4 d at 30 °C or 2 d at 37 °C or 39.5 °C, cells were rinsed with 0.15 M NaCl, fixed with methanol, and stained with a solution of 0.2% toluidine blue, 0.2% azure blue, and 1% boric acid. The EOP was calculated by dividing titers at 39.5 °C by the titer at 30 °C. We found that all of the ts mutants produced the same titer at 30 °C as at 33 °C and in all cases ± 39 °C was non-permissive for plaque formation of the ts mutants.
Complementation analysis.
Classic or genetic complementation was done by infecting 17Cl-1cells in 35 mm petri dishes either singly with each mutant or doubly with two mutants at a multiplicity of infection (MOI) of 20–30 pfu/cell. After incubation at room temperature or 30 °C for 30 min, the virus inoculum was removed and the infected cells were rinsed with HBSS and re-fed with low pH DMEM or DMEM supplemented with 6% FBS, penicillin (100 units/ml), and streptomycin (100g/ml). The infected cells were incubated at 39.5 °C or 40 °C in 10% CO2. One hour later the cells were rinsed again and re-fed with warmed medium and the dishes returned to the incubator until 8 hpi. The medium was harvested, clarified at 10,000 rpm for 5 min, and virus titers were determined by plaque assays at 30 °C. CIs were calculated using the following formula:
A CI greater than two between mutant pairs was consistent with complementation, i.e., ± 4-fold difference above background, while a CI less than two was negative for complementation [43].
Biochemical complementation was done by mock-infecting or infecting 17Cl-1cells at 30 pfu/cell with MHV-A59, or one of the ts mutants, or with a mix of 15 pfu/cell each of two ts mutants. After the adsorption period at room temperature for 30 min, the virus inoculum was removed, 1 ml of prewarmed medium containing dactinomycin and 3H-uridine (1.85 MBq/ml) was added and the cells were incubated at 40 °C. The wt and the ts mutant-infected cells were harvested at 8 hpi and the 3H-uridine incorporation into trichloroacetic acid-precipitated RNA was determined.
Viral RNA synthesis.
Viral RNA synthesis was measured by determining the amount of 3H-uridine incorporated in the presence of dactinomycin (20 μg/ml) into acid-precipitable material. [5-3H] uridine (≥1.0 TBq/mmol) was added to the medium at either 1.85 or 7.4 MBq/ml. After incubation, the radioactive medium was removed and the cells dissolved with 5% lithium dodecyl sulfate and 200 μg/ml proteinase K in LEH buffer (0.1 M LiCl, 0.001 M EDTA, 0.01 M HEPES, [pH 6.6]) at 2–5 × 105 cells per ml. The DNA was sheared by repeated passage through a 27-gauge needle attached to a 1-ml tuberculin syringe. Triplicate samples of 5 × 104 cells were precipitated with trichloroacetic acid and the precipitates collected on glass fiber filters (Whatman Incorporated, Clifton, New Jersey, United States), dried under a heat lamp, and the radioactivity determined by liquid scintillation spectroscopy. To measure negative-strand synthesis, the dissolved cells were extracted with low pH phenol (pH 4.3), which removed DNA from the aqueous phase, and then with cholorofom:isoamyl alcohol (95:5), and the RNA was precipitated with ethanol. RF RNA was generated by treatment of the RNA with RNase T1 (1U/ug RNA, 30 °C for 30 min in 0.3 M NaCl) and collected by chromatography on CF-11 cellulose and ethanol precipitation. The incorporation of 3H-uridine into negative strands was measured by denaturing the RF RNA with heat and annealing in the presence of >100-fold excess of unlabeled RNA obtained from purified virions of MHV [24].
Isolation of viral RNA.
Two different procedures were used to obtain viral RNA for RT-PCR and sequencing. Virions were purified from ~3 × 108 17Cl-1 cells that had been infected at a MOI of ~10 pfu/cell and incubated in low pH DMEM at 30–33 °C. The medium from the infected cells (~225 ml) was collected at 24 hpi and clarified at 4,000 rpm for 30 min. The virions were pelleted by centrifugation at 24,000 rpm for 3 h at 4 °C. The virus pellet was allowed to suspend overnight on ice in 0.4 ml/tube of 0.15M NaCl and 20 mM HEPES (pH 6.6). The suspended virus from six tubes was pooled and layered over one SW28 tube containing a linear gradient of 40% potassium tartrate (bottom) and 20% glycerol (top), in 0.0.2 M HEPES (pH 7.4). After centrifugation at 24,000 rpm for 3–4 h at 4 °C, the visible band containing the virions was collected, diluted, and pelleted by centrifugation at 24,000 rpm for 3 h at 4 °C. The pelleted virions were suspended in 0.15M NaCl and 20 mM HEPES (pH 6.6), and LiDS and proteinase K were added to 5% and 400 μg/ml, respectively, After incubation at 42 °C for 10 min, the viral RNA was extracted with phenol followed by chloroform:isoamylalcohol (19:1). Viral RNA was ethanol-precipitated and the pellet was washed with 70% ethanol, dried under vacuum, and resuspended in water. Alternatively, 107 17Cl-1 cells were infected with virus, incubated for 13 h at 30 °C to 33 °C in 7.5% CO2. The poly(A)-containing RNA was then isolated from the infected cells using oligo-dT25 Dynabeads as described previously [62].
RT-PCR and sequencing.
The entire replicase gene-coding region (ORF1a and ORF1b) was sequenced for eight ts mutant and revertant pairs. To do this, we used a set of 121 synthetic oligonucleotides that are complementary to sequences spaced at approximately 350 nucleotide intervals along the positive- and negative-strand copies of the viral RNA (sequences available on request). Five oligonucleotides, P17, P31, P46, P61, and P65, were used to prime the RT of viral RNA with Superscript II RT (Invitrogen, Carlsbad, California, United States). The reaction mix (20 μl), which contained, in addition to pre-supplied buffer, 35 ng of primer, 10–100 ng of viral RNA, 1 mM dNTPs, 10 mM DTT, 25 U of RNAguard (Amersham, Little Chalfont, United Kingdom), and 200 U of reverse transcriptase, was incubated at 42 °C for 60 min and then at 94 °C for 2 min. The five cDNA templates were then amplified using eight primer pairs, P1/P16, P2/P22, P3/P30, P4/P38, P5/P45, P6/P53, P7/P60, and P8/P64, and thermostable, recombinant Taq DNA polymerase. The reaction mix (100 μl), which contained, in addition to pre-supplied buffer, 70 ng of primer pair, 1 μl of RT reaction product, 200 μM dNTPs, 2 mM MgCl2, and 2.5 U of DNA polymerase, was incubated at 94 °C for 1 min, then 94 °C for 20 s, 50 °C for 20 s, 68 °C for 3 min, for a total of 35 cycles and a final 10-min extension at 68 °C. The PCR reaction products were purified by ethanol precipitation using ammonium acetate. Finally, sequence analysis was done using primers P1–P121 and standard cycle sequencing methods. Sequencing reaction mixes (10 μl), which contained 70 ng of primer, 100 ng of PCR product, and 3 μl of cycle sequencing mix (BigDye Terminator v.3.1, Applied Biosystems, Foster City, California, United States), were incubated at 96 °C for 10 s, 50 °C for 5 s, and 60 °C for 4 min, for a total of 25 cycles. The reaction products were purified by retention on a size exclusion membrane (Montage™ SEQ96, Millipore, Billerica, Massachusetts, United States) as described by the manufacturer; eluted and analyzed with an ABI 310 Prism Genetic Analyzer. Computer-assisted analysis of sequence data was done using the Lasergene bio-computing software (DNASTAR).
Supporting Information
Figure S1 Plaque Morphology of Alb ts17 Revertants
The plaque morphologies of Alb ts17L and Alb ts17S are illustrated. Alb ts17 had a reversion (back mutation) frequency of 2 × 10−6 and there was a mixture of large and small plaques at 40 °C. The virus from the small and large plaques produced progeny that formed uniformly small or large plaques at 40 °C, respectively. At 30 °C, both 17RL and 17RS produced plaques of equal diameter and Alb 17RL produced the same size plaques at 40 °C as the parental or wt MHV-A59.
(1.7 MB PPT)
Click here for additional data file.
Table S1 Phenotypic Analysis of MHV-A59 ts Mutants
(31 KB DOC)
Click here for additional data file.
We would like to thank Barbara Schelle and Tamara Jones for technical help. This work was supported by grants from the German Research Council and Wellcome Trust (S. G. Siddell) and the National Institutes of Health (S. G. Sawicki and D. L. Sawicki).
Competing interests. The authors have declared that no competing interests exist.
Author contributions. S. G. Sawicki, D. L. Sawicki, and S. G. Siddell conceived and designed the experiments. S. G. Sawicki, D. L. Sawicki, D. Younker, Y. Meyer, V. Thiel, H. Stokes, and S. G. Siddell performed the experiments and analyzed the data. S. G. Sawicki, V. Thiel, and S. G. Siddell wrote the paper.
Abbreviations
CHcycloheximide
CIcomplementation index
EOPefficiency of plating
hpihour post-infection
ORFopen reading frame
pfuplaque-forming unit
TRStranscription regulating sequence
tstemperature-sensitive
wtwild-type
==== Refs
References
Siddell SG Ziebuhr J Snijder EJ 2005 Coronaviruses, toroviruses, and arteriviruses Mahy BWJ ter Meulen V Virology. 10th ed London Hodder Arnold 823 855
Woo PC Lau SK Chu CM Chan KH Tsoi HW 2005 Characterization and complete genome sequence of a novel coronavirus, coronavirus HKU1, from patients with pneumonia J Virol 79 884 895 15613317
van der Hoek L Pyrc K Jebbink MF Vermeulen-Oost W Berkhout RJ 2004 Identification of a new human coronavirus Nat Med 10 368 373 15034574
Stadler K Masignani V Eickmann M Becker S Abrignani S 2003 SARS—beginning to understand a new virus Nat Rev Microbiol 1 209 218 15035025
Sawicki SG Sawicki DL 1995 Coronaviruses use discontinuous extension for synthesis of subgenome-length negative strands Adv Exp Med Biol 380 499 506 8830530
Pasternak AO Spaan WJ Snijder EJ 2004 Regulation of relative abundance of arterivirus subgenomic mRNAs J Virol 78 8102 8113 15254182
Curtis KM Yount B Sims AC Baric RS 2004 Reverse genetic analysis of the transcription regulatory sequence of the coronavirus transmissible gastroenteritis virus J Virol 78 6061 6066 15141005
Bredenbeek PJ Pachuk CJ Noten AF Charite J Luytjes W 1990 The primary structure and expression of the second open reading frame of the polymerase gene of the coronavirus MHV-A59; a highly conserved polymerase is expressed by an efficient ribosomal frameshifting mechanism Nucleic Acids Res 18 1825 1832 2159623
Brockway SM Clay CT Lu XT Denison MR 2003 Characterization of the expression, intracellular localization, and replication complex association of the putative mouse hepatitis virus RNA-dependent RNA polymerase J Virol 77 10515 10527 12970436
Ziebuhr J Snijder EJ Gorbalenya AE 2000 Virus-encoded proteinases and proteolytic processing in the Nidovirales J Gen Virol 81 853 879 10725411
Snijder EJ Bredenbeek PJ Dobbe JC Thiel V Ziebuhr J 2003 Unique and conserved features of genome and proteome of SARS-coronavirus, an early split-off from the coronavirus group 2 lineage J Mol Biol 331 991 1004 12927536
Baker SC Yokomori K Dong S Carlisle R Gorbalenya AE 1993 Identification of the catalytic sites of a papain-like cysteine proteinase of murine coronavirus J Virol 67 6056 6063 8396668
Lu Y Lu X Denison MR 1995 Identification and characterization of a serine-like proteinase of the murine coronavirus MHV-A59 J Virol 69 3554 3559 7745703
Bhardwaj K Guarino L Kao CC 2004 The severe acute respiratory syndrome coronavirus Nsp15 protein is an endoribonuclease that prefers manganese as a cofactor J Virol 78 12218 12224 15507608
Putics A Filipowicz W Hall J Gorbalenya AE Ziebuhr J 2005 ADP-ribose-1"-monophosphatase: A conserved coronavirus enzyme that is dispensable for viral replication in tissue culture J Virol 79 12721 12731 16188975
Ivanov KA Hertzig T Rozanov M Bayer S Thiel V 2004 Major genetic marker of nidoviruses encodes a replicative endoribonuclease Proc Natl Acad Sci U S A 101 12694 12699 15304651
Ivanov KA Thiel V Dobbe JC van der Meer Y Snijder EJ 2004 Multiple enzymatic activities associated with severe acute respiratory syndrome coronavirus helicase J Virol 78 5619 5632 15140959
Ivanov KA Ziebuhr J 2004 Human coronavirus 229E nonstructural protein 13: Characterization of duplex-unwinding, nucleoside triphosphatase, and RNA 5′-triphosphatase activities J Virol 78 7833 7838 15220459
Cheng A Zhang W Xie Y Jiang W Arnold E 2005 Expression, purification, and characterization of SARS coronavirus RNA polymerase Virology 335 165 176 15840516
Seybert A Posthuma CC van Dinten LC Snijder EJ Gorbalenya AE 2005 A complex zinc finger controls the enzymatic activities of nidovirus helicases J Virol 79 696 704 15613297
Sutton G Fry E Carter L Sainsbury S Walter T 2004 The nsp9 replicase protein of SARS-coronavirus, structure, and functional insights Structure (Camb) 12 341 353 14962394
Egloff MP Ferron F Campanacci V Longhi S Rancurel C 2004 The severe acute respiratory syndrome-coronavirus replicative protein nsp9 is a single-stranded RNA-binding subunit unique in the RNA virus world Proc Natl Acad Sci U S A 101 3792 3796 15007178
Yang H Yang M Ding Y Liu Y Lou Z 2003 The crystal structures of severe acute respiratory syndrome virus main protease and its complex with an inhibitor Proc Natl Acad Sci U S A 100 13190 13195 14585926
Sawicki SG Sawicki DL 1986 Coronavirus minus-strand RNA synthesis and effect of cycloheximide on coronavirus RNA synthesis J Virol 57 328 334 2867230
Sawicki SG Sawicki DL 1990 Coronavirus transcription: Subgenomic mouse hepatitis virus replicative intermediates function in RNA synthesis J Virol 64 1050 1056 2154591
Sawicki D Wang T Sawicki S 2001 The RNA structures engaged in replication and transcription of the A59 strain of mouse hepatitis virus J Gen Virol 82 385 396 11161278
Wang T Sawicki SG 2001 Mouse hepatitis virus minus-strand templates are unstable and turn over during viral replication Adv Exp Med Biol 494 491 497 11774513
Kim YN Jeong YS Makino S 1993 Analysis of cis-acting sequences essential for coronavirus defective interfering RNA replication Virology 197 53 63 8212595
Goebel SJ Hsue B Dombrowski TF Masters PS 2004 Characterization of the RNA components of a putative molecular switch in the 3′ untranslated region of the murine coronavirus genome J Virol 78 669 682 14694098
Pasternak AO van den Born E Spaan WJ Snijder EJ 2003 The stability of the duplex between sense and antisense transcription-regulating sequences is a crucial factor in arterivirus subgenomic mRNA synthesis J Virol 77 1175 1183 12502834
Zuniga S Sola I Alonso S Enjuanes L 2004 Sequence motifs involved in the regulation of discontinuous coronavirus subgenomic RNA synthesis J Virol 78 980 994 14694129
van Marle G Luytjes W van der Most RG van der Straaten T Spaan WJ 1995 Regulation of coronavirus mRNA transcription J Virol 69 7851 7856 7494297
Zhang X Liu R 2000 Identification of a non-canonical signal for transcription of a novel subgenomic mRNA of mouse hepatitis virus: Implication for the mechanism of coronavirus RNA transcription Virology 278 75 85 11112483
Huang P Lai MM 1999 Polypyrimidine tract-binding protein binds to the complementary strand of the mouse hepatitis virus 3′ untranslated region, thereby altering RNA conformation J Virol 73 9110 9116 10516017
Huang P Lai MM 2001 Heterogeneous nuclear ribonucleoprotein a1 binds to the 3′-untranslated region and mediates potential 5′-3′-end cross talks of mouse hepatitis virus RNA J Virol 75 5009 5017 11333880
Nanda SK Leibowitz JL 2001 Mitochondrial aconitase binds to the 3′ untranslated region of the mouse hepatitis virus genome J Virol 75 3352 3362 11238861
Shen X Masters PS 2001 Evaluation of the role of heterogeneous nuclear ribonucleoprotein A1 as a host factor in murine coronavirus discontinuous transcription and genome replication Proc Natl Acad Sci U S A 98 2717 2722 11226306
Diamond SE Kirkegaard K 1994 Clustered charged-to-alanine mutagenesis of poliovirus RNA-dependent RNA polymerase yields multiple temperature-sensitive mutants defective in RNA synthesis J Virol 68 863 876 8289389
Schaad MC Stohlman SA Egbert J Lum K Fu K 1990 Genetics of mouse hepatitis virus transcription: Identification of cistrons which may function in positive and negative strand RNA synthesis Virology 177 634 645 2164727
Sturman LS Eastwood C Frana MF Duchala C Baker F 1987 Temperature-sensitive mutants of MHV-A59 Adv Exp Med Biol 218 159 168 2829525
Koolen MJ Osterhaus AD Van Steenis G Horzinek MC Van der Zeijst BA 1983 Temperature-sensitive mutants of mouse hepatitis virus strain A59: Isolation, characterization, and neuropathogenic properties Virology 125 393 402 6301146
Malpica JM Fraile A Moreno I Obies CI Drake JW 2002 The rate and character of spontaneous mutation in an RNA virus Genetics 162 1505 1511 12524327
Burge BW Pfefferkorn ER 1966 Complementation between temperature-sensitive mutants of Sindbis virus Virology 30 214 223 5919229
Stalcup RP Baric RS Leibowitz JL 1998 Genetic complementation among three panels of mouse hepatitis virus gene 1 mutants Virology 241 112 121 9454722
Coley SE Lavi E Sawicki SG Fu L Schelle B 2005 Recombinant mouse hepatitis virus strain A59 from cloned, full-length cDNA replicates to high titers in vitro and is fully pathogenic in vivo J Virol 79 3097 3106 15709029
Henikoff S Henikoff JG 1992 Amino acid substitution matrices from protein blocks Proc Natl Acad Sci U S A 89 10915 10919 1438297
Anand K Palm GJ Mesters JR Siddell SG Ziebuhr J 2002 Structure of coronavirus main proteinase reveals combination of a chymotrypsin fold with an extra alpha-helical domain Embo J 21 3213 3224 12093723
Anand K Ziebuhr J Wadhwani P Mesters JR Hilgenfeld R 2003 Coronavirus main proteinase (3CLpro) structure: Basis for design of anti-SARS drugs Science 300 1763 1767 12746549
Rost B 1996 PHD: Predicting one-dimensional protein structure by profile-based neural networks Methods Enzymol 266 525 539 8743704
De I Sawicki SG Sawicki DL 1996 Sindbis virus RNA-negative mutants that fail to convert from minus-strand to plus-strand synthesis: Role of the nsP2 protein J Virol 70 2706 2719 8627744
Tijms MA van Dinten LC Gorbalenya AE Snijder EJ 2001 A zinc finger-containing papain-like protease couples subgenomic mRNA synthesis to genome translation in a positive-stranded RNA virus Proc Natl Acad Sci U S A 98 1889 1894 11172046
Graham RL Sims AC Brockway SM Baric RS Denison MR 2005 The nsp2 replicase proteins of murine hepatitis virus and severe acute respiratory syndrome coronavirus are dispensible for virus replication J Virol 79 13399 13411 16227261
Ivanov KA Hertzig T Rozanov M Bayer S Thiel V 2004 Major genetic marker of nidoviruses encodes a replicative endoribonuclease Proc Natl Acad Sci U S A 101 12694 12699 15304651
Sawicki SG Sawicki DL 1986 The effect of loss of regulation of minus-strand RNA synthesis on Sindbis virus replication Virology 151 339 349 3705466
Sawicki SG Sawicki DL 1986 The effect of overproduction of nonstructural proteins on alphavirus plus-strand and minus-strand RNA synthesis Virology 152 507 512 3014730
Schiller JJ Kanjanahaluethai A Baker SC 1998 Processing of the coronavirus MHV-JHM polymerase polyprotein: Identification of precursors and proteolytic products spanning 400 kilodaltons of ORF1a Virology 242 288 302 9514967
Lu XT Sims AC Denison MR 1998 Mouse hepatitis virus 3C-like protease cleaves a 22-kilodalton protein from the open reading frame 1a polyprotein in virus-infected cells and in vitro J Virol 72 2265 2271 9499085
Denison MR Spaan WJ van der Meer Y Gibson CA Sims AC 1999 The putative helicase of the coronavirus mouse hepatitis virus is processed from the replicase gene polyprotein and localizes in complexes that are active in viral RNA synthesis J Virol 73 6862 6871 10400784
Ziebuhr J Herold J Siddell SG 1995 Characterization of a human coronavirus (strain 229E) 3C-like proteinase activity J Virol 69 4331 4338 7769694
Sturman LS Takemoto KK 1972 Enhanced growth of a murine coronavirus in transformed mouse cells Infect Immun 6 501 507 4564284
Weiland E Mussgay M Weiland F 1978 Nonproducer malignant tumor cells with rescuable sarcoma virus genome isolated from a recurrent Moloney sarcoma J Exp Med 148 408 423 151722
Thiel V Rashtchian A Herold J Schuster DM Guan N 1997 Effective amplification of 20-kb DNA by reverse transcription PCR Anal Biochem 252 62 70 9324942
|
16341254
|
PMC1298938
|
CC BY
|
2021-01-05 12:11:25
|
no
|
PLoS Pathog. 2005 Dec 9; 1(4):e39
|
utf-8
|
PLoS Pathog
| 2,005 |
10.1371/journal.ppat.0010039
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1631841210.1371/journal.pmed.0030002Research ArticleAllergy/ImmunologyRheumatologyRheumatologyImmunology and AllergyAutoimmune DiseasesAntibodies against Human Cytomegalovirus in the Pathogenesis of Systemic Sclerosis: A Gene Array Approach Anti-hCMV Antibodies and FibroblastsLunardi Claudio
1
*Dolcino Marzia
2
Peterlana Dimitri
1
Bason Caterina
1
Navone Riccardo
1
Tamassia Nicola
3
Beri Ruggero
1
Corrocher Roberto
1
Puccetti Antonio
2
4
1Department of Clinical and Experimental Medicine, Section of Internal Medicine, University of Verona, Verona, Italy2Institute Giannina Gaslini, Genova, Italy3Department of Pathology, Section of General Pathology, University of Verona, Verona, Italy4Department of Experimental Medicine, Section of Histology, University of Genova, Genova, ItalyHuizinga Tom Academic EditorLeiden University Medical CenterNetherlands*To whom correspondence should be addressed. E-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
Author Contributions: CL and AP conceived the idea and were the principal investigators of the Genova and Verona teams, respectively. MD and DP performed the analysis of all the genes up- and downregulated. CB was responsible for cell cultures and purification of the antibodies. RN performed the Western blotting and helped in the analysis of the genes. NT performed the real-time quantitative PCR. RB measured the soluble molecules in the patients' and controls' sera. RC advised on the preparation of the report. CL and AP wrote the article, with input from DP.
1 2006 6 12 2005 3 1 e27 3 2005 22 9 2005 Copyright: © 2006 Lunardi et al.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
CMV and Triggers of Systemic Sclerosis
Background
Systemic sclerosis is an autoimmune disease characterized by immunological abnormalities, vascular damage, and fibroblast proliferation. We have previously shown that a molecular mimicry mechanism links antibodies against the human-cytomegalovirus-derived protein UL94 to the pathogenesis of systemic sclerosis. The UL94 epitope shows homology with NAG-2, a surface molecule highly expressed on endothelial cells. Anti-UL94 peptide antibodies purified from patients' sera induce apoptosis of endothelial cells upon engagement of the NAG-2–integrin complex.
Methods and Findings
We show here that NAG-2 is expressed on dermal fibroblasts and that anti-UL94 antibodies bind to fibroblasts. We have used the gene array strategy (Affimetrix oligonucleotide microarrays) to analyze the transcriptional profile in response to a 4-h and an 8-h treatment with antibodies against the UL94 peptide in endothelial cells and dermal fibroblasts. Exposure of endothelial cells to anti-UL94 antibodies had a profound impact on gene expression, resulting in the upregulation of 1,645 transcripts. Several gene clusters were upregulated including genes encoding adhesion molecules, chemokines, colony-stimulating factors (CSFs), growth factors, and molecules involved in apoptosis. Following antibody stimulation, dermal fibroblasts showed an upregulation of 989 transcripts and acquired a “scleroderma-like” phenotype. Indeed, genes involved in extracellular matrix deposition, growth factors, chemokines, and cytokines were upregulated. We confirmed the microarray results by real-time quantitative polymerase chain reaction and by measuring some of the corresponding proteins with ELISA and Western blotting.
Conclusion
Our results show that anti-human-cytomegalovirus antibodies may be linked to the pathogenesis of systemic sclerosis not only by inducing endothelial cell activation and apoptosis but also by causing activation of fibroblasts, one of the hallmarks of the disease.
Anti-human-cytomegalovirus antibodies may be linked to the pathogenesis of systemic sclerosis by inducing endothelial cell activation and apoptosis, and fibroblast activation and proliferation.
==== Body
Introduction
Systemic sclerosis (SSc) is an autoimmune disease characterized by three main features: (i) structural and functional vascular abnormalities with perivascular infiltration of mononuclear inflammatory cells, intimal proliferation, and luminal narrowing at both the arteriolar and arterial levels, (ii) immunologic abnormalities, both humoral and cellular, such as the presence of autoantibodies to intracellular and cell surface antigens, and perivascular T cell infiltration of the skin and internal organs, and (iii) excessive extracellular matrix deposition, leading to fibrosis of the skin and of internal organs [1].
Autoantibodies directed against intracellular antigens are associated with SSc and differentiate two distinct clinical subsets: anticentromere antibodies are found in SSc with limited cutaneous involvement, while anti–DNA topoisomerase I antibodies are associated with SSc with diffuse cutaneous involvement [2]. Moreover, autoantibodies directed against cell surface antigens might induce endothelial cell damage and apoptosis, considered a primary event in the pathogenesis of the disease [3,4].
Latent human cytomegalovirus (hCMV) infection may contribute to progression of SSc through its ability to infect endothelial cells [5]. Indirect evidence for the association between hCMV and SSc comes from the prevalence of anti-hCMV antibodies in patients affected by the disease [6]. Moreover, monoclonal antibodies against topoisomerase I were found to recognize a pentapeptide of the autoantigen sharing homology with the hCMV-derived UL70 protein, suggesting the activation of autoreactive B cell clones by a molecular mimicry mechanism [7]. Furthermore, some patients with chronic graft-versus-host disease develop SSc-like lesions with the presence of typical autoantibodies such as anti–topoisomerase I [5], and hCMV infection is associated with an increased risk for the development of chronic graft-versus-host disease [8]. Finally, murine sclerodermatous graft-versus-host disease is one of the animal models for human scleroderma [9,10].
In a previous study we provided direct evidence for a molecular mimicry mechanism by which antibodies against a hCMV-derived protein can be linked to endothelial cell damage in patients with SSc [11]. In the majority of patients' sera there are antibodies directed against an epitope (VTL
GGAGIWLPP) contained within UL94, a hCMV-derived protein expressed in infected cells with very late kinetics. UL94 is localized in the nucleus of infected cells and may be involved in the regulation of viral and/or cellular gene expression. The UL94 epitope shows homology with NAG-2 [12], a cell surface molecule highly expressed on non-stressed endothelial cells and associated with integrins. Affinity purified anti-UL94 peptide IgG antibodies recognize NAG-2 in a whole cell lysate and induce apoptosis of non-stressed endothelial cells upon engagement of the NAG-2–integrin complex [11]. Therefore, we propose that hCMV is linked to the pathogenesis of SSc through a particular subset of anti-hCMV antibodies that specifically interacts with a normally expressed endothelial cell surface receptor sharing similarity with the UL94 viral protein. The engagement of the receptor results in endothelial cell apoptosis, considered the primary pathogenic event in SSc.
Another fundamental feature of SSc is the fibrosis of the skin and internal organs because of increased extracellular matrix deposition [13]. Indeed, fibroblasts are thought to play a major role in the pathogenesis of the disease. They are directly involved in the synthesis of many extracelluar matrix components, and the dysregulation of extracellular matrix turnover is central to fibrosis development in SSc. Scleroderma fibroblasts display a variety of phenotypic defects that range from increased synthesis of multiple matrix proteins to abnormalities of cell surface receptors and signaling pathways [14].
While a direct link between endothelial cell damage in SSc and hCMV infection has been shown, a correlation between hCMV and fibrosis is still lacking. In the present study we wanted to verify whether the NAG-2 receptor is expressed also on normal fibroblasts and whether the anti-hCMV antibodies bind normal dermal fibroblasts upon interaction with the NAG-2 receptor. Moreover, we decided to use a DNA microarray approach to analyze the effects of these antibodies on endothelial cells and fibroblasts, in order to better understand the role played by hCMV in the pathogenesis of the disease. Our findings at gene level were confirmed with real-time quantitative polymerase chain reaction and with the analysis of a panel of proteins both in the supernatants of the cells exposed to the antibodies and in the sera of patients affected by SSc.
Methods
Patients and Controls
Eighty-one SSc patients (75 women and six men, mean ± standard deviation age 54.5 ± 14.1 y) from northern Italy were studied. All patients fulfilled the American College of Rheumatology criteria for SSc [15]. Sixty of them were affected by limited cutaneous SSc and 21 by diffuse cutaneous SSc. Sixty sex- and age-matched individuals without SSc were used as controls. Blood was obtained from all the patients and controls after informed written or oral consent.
Peptide Synthesis
The UL94-derived peptide (VTL
GGAGIWLLP) and the irrelevant control peptide (VTLPKDSDVELP)—chosen on the basis of its absence of relevant homology with the viral peptide, with the NAG-2 molecule, and with human self-antigens according to BLASTP via the NCBI BLAST network service [11]—were synthesized by solid phase synthesis with the 9-fluorenylmethoxycarbonyl strategy [16].
Purification of Anti-Peptide Antibodies
To affinity purify IgG antibodies against either the UL94 peptide or the control peptide, each peptide (5 mg/g of dried sepharose powder) was coupled to sepharose 4B (Pharmacia, Uppsala, Sweden), according to the manufacturer's instructions. Pooled sera from ten different patients with SSc were diluted in PBS. Aliquots of the diluted serum samples were applied to each column separately. The columns were washed with PBS. Bound immunoglobulins were eluted with 0.1 M glycine (pH 2.5) and dialyzed against PBS.
Cell Culture
Human endothelial cells (HUVECs) and human dermal fibroblast as well as their growth media were purchased from Promocell Bioscience Alive (Heidelberg, Germany). HUVECs were used between passages 2 and 5, while dermal fibroblasts between passages 3 and 6. Cell monolayers were stimulated with antibodies affinity purified against the UL94 peptide or the irrelevant control peptide (20 μg/ml) for various intervals of time as described in the text. Cell pellets were used for gene array experiments, while soluble mediators were measured in the cell supernatants.
FACS Analysis
For FACS (fluorescence-activated cell sorting) analysis, cells were incubated with specific or control antibodies (5 μg/ml) for 30 min on ice. Antibody binding was revealed with fluorescein-isothiocyanate-conjugated anti-human IgG antibodies (SouthernBiotech, Birmingham, Alabama, United States). Samples were run on a FACScan flow cytometer (Becton Dickinson, Franklin Lakes, New Jersey, United States).
Western Blot Analysis
Human dermal fibroblasts were lysed in cold lysis buffer (0.5% NP-40, 10 mM TRIS [pH 7.4], 0.15 M sodium chloride, 5 mM magnesium chloride), and lysates were immunoprecipitated with rabbit anti-NAG-2 antibodies [11] cross-linked to sepharose 4B (Pharmacia). Eluted proteins were resolved by SDS-PAGE and transferred to nitrocellulose membrane (Amersham Bioscience, Piscataway, New Jersey, United States). Blots were incubated with either anti-NAG-2 antibodies or human affinity purified anti-peptide antibodies (10 μg/ml). The Renaissance Chemiluminescence Kit (NEN, Boston, Massachusetts, United States) was used for detection.
Competitive Enzyme-Linked Immunosorbent Assay
Enzyme-linked immunosorbent assays (ELISAs) were carried out as previously described [11,17]. Briefly, for the binding of antibodies to fibroblasts, affinity purified anti-hCMV peptide antibodies (10 μg/ml) were diluted in PBS containing 1% BSA and were incubated for 1 h at 37 °C with fibroblasts that had been previously fixed with 0.1% glutaraldehyde. Bound IgG was detected by a further 1 h incubation with a peroxidase-conjugated antibody against human IgG (Amersham Bioscience). For the competitive assay, the amount of antibody that gave 50% of the maximum binding to fibroblast was preincubated for 1 h at 37 °C with different amounts of competitors or buffer and then transferred to the fibroblast-coated plates. The assay was then carried out as the direct binding assay.
Proliferation Assay
To assess cell proliferation, fibroblasts (5000 cells/well) were cultured for various intervals of time in microtiter plates in the presence or absence of antibodies (15 μg/ml affinity purified antibodies). Cell viability was assessed using the commercially available kit (Alexis Biochemicals, San Diego, California, United States).
Preparation of cRNA and Array Analysis
Preparation of cRNA, hybridization, and scanning of probe arrays were performed according to the protocols of the manufacturer (Affymetrix, Santa Clara, California, United States) by the Genopolis Consortium (University of Milano-Bicocca, Italy) using the Human Genome U133A GeneChip (Affymetrix). The Human Genome U133A GeneChip is a single array representing 14,500 well-characterized human genes and including more than 22,000 probe sets and 500,000 distinct oligonucleotide features.
The different gene expression patterns were analyzed by using Array Assist version 2.0 (Stratagene, La Jolla, California, United States), which calculated a robust multi-array average of background-adjusted, normalized, and log-transformed intensity values applying the robust multi-array average algorithm [18,19]. With this software, the mean optical background level for each array was subtracted from the signal intensity for each probe. Finally, the normalized, background-corrected data were transformed to the log2 scale. A signal log2 ratio of 1.0 indicates an increase of the transcript level by 2-fold change (2 F.C.) and −1.0 indicates a decrease by 2-fold (−2 F.C.). A signal log2 ratio of zero would indicate no change.
In our study, we analyzed the gene expression profiles in both endothelial cells and fibroblasts stimulated with antibodies affinity purified against the UL94 peptide (test samples) or with antibodies affinity purified against an irrelevant peptide (control samples) for 4 and 8 h. Genes were selected for final consideration when their expression (F.C.) was at least 2-fold different in the test sample versus control sample at at least one time point. The microarray results have been reported according to the MIAME guidelines and deposited in the public repository ArrayExpress (http://www.ebi.ac.uk/arrayexpress; accession number E-MEXP-382).
Quantitative Real-Time Polymerase Chain Reaction
Total RNA was extracted from endothelial cells and fibroblasts using TRIzol reagent (Invitrogen, Carlsbad, California, United States), following manufacturer's instructions. One microgram of total RNA from each sample was treated with amplification grade DNase I and then used as a template for the reverse transcription reaction, using random hexamers and SuperScript II Reverse Transcriptase (Invitrogen). All samples were reverse transcribed under the same conditions and from the same reverse transcription master mix, in order to minimize differences in reverse transcription efficiency. Triplicate quantitative real-time polymerase chain reactions (Q-PCRs) for each sample were performed in 20 μl containing 20 ng of cDNA, 10 μl of 2× Platinum SYBR Green qPCR SuperMix UDG (Invitrogen), and primers at a final concentration of 200 nM. Table 1 shows the primers used. The Q-PCR reactions were performed in 96-well plates with optical caps using the DNA Engine Opticon 2 System (MJ Research, Waltham, Massachusetts, United States). The reaction conditions were identical for all primer sets as follows: 50 °C for 2 min, 95 °C for 2 min, and then 40 cycles of 95 °C for 15 s and 60 °C for 1 min. Control wells containing no template were used to exclude the presence of contaminating template molecules and to identify potential primer–dimer products from the dissociation curve analysis. Amplification plots were analyzed using Opticon Monitor version 2.02 (MJ Research), and data were calculated with QGene (http://www.qgene.org/) and expressed as mean normalized expression. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was selected as a normalizing gene according to its stable expression levels.
Table 1 Primers Used for Q-PCR
Cytokines, Chemokines, and Adhesion Molecules
Aliquots of sera and supernatants were frozen at −80 °C until assayed. The soluble mediators were measured with commercially available ELISAs according to the manufacturer's protocol. Measurement below detectable levels was used as the lower cut-off limit of the assay, according to the instructions of the manufacturer. The value recorded was the mean of two measurements. ELISA kits for Vascular endothelial growth factor (VEGF), Interleukin-6 (IL-6), IL-8, Transforming growth factor beta 1 (TGF-beta 1), and Monocyte chemotactic protein 1 (MCP-1) were purchased from Amersham Bioscience; IL-11, MCP-3, soluble Intercellular adhesion molecule 1 (ICAM-1), soluble E-selectin, and Vascular cell adhesion molecule 1 (VCAM-1) from R&D Systems (Minneapolis, Minnesota, United States); Interferon-gamma-inducible protein 10 (IP10) from Bender MedSystems (Vienna, Austria); Endothelin 1 (ET-1) from Assay Designs (Ann Arbor, Michigan, United States); C-terminal Propeptide of Collagen type I (Pro-Col I) from Quidel Corporation (San Diego, California, United States); and Epidermal growth factor (EGF) from Chemicon International (Temecula, California, United States).
Gene Ontology Analysis
We performed a Gene Ontology (GO) analysis using Array Assist version 2.0 (Stratagene).
Statistical Analysis
Statistical testing was performed using StatsDirect (StatsDirect, Cheshire, United Kingdom). The significance of differences between patients and controls was determined using the unpaired Student's t-test; p < 0.05 was considered statistically significant. For sake of clearness the values are expressed as mean with 95% confidence interval.
Results
Anti-hCMV Peptide Antibodies Bind to Normal Dermal Fibroblasts
To verify whether anti-hCMV peptide antibodies bind to human dermal fibroblasts, we performed a FACS analysis using affinity purified anti-UL94 peptide IgG antibodies and dermal fibroblasts. As shown in Figure1A and 1B, anti-peptide antibodies were able to bind dermal fibroblasts. We also showed that the NAG-2 receptor is expressed on the surface of dermal fibroblasts and that this molecule is recognized by anti-hCMV peptide antibodies (Figure 1C). The specificity of the interaction of such antibodies with the NAG-2 receptor was further confirmed by a competitive ELISA that demonstrated that the viral peptide could displace the binding of the antibodies from the surface of the dermal fibroblasts (Figure 1D). Since we have previously demonstrated that anti-hCMV antibodies are able to inhibit cell proliferation and induce apoptosis of endothelial cells upon engagement of the NAG-2 molecule [11], we next wanted to evaluate whether the interaction between the antibodies and dermal fibroblasts had some functional consequence for these cells. As shown in Figure 1E, dermal fibroblasts incubated with the anti-hCMV antibodies proliferated normally, and after 24 and 48 h the rate of proliferation was slightly higher than for cells incubated with antibodies against an irrelevant peptide. These data indicate that anti-hCMV peptide antibodies recognize NAG-2 expressed on dermal fibroblasts and that this interaction does not inhibit cell proliferation of the target cells.
Figure 1 Anti-hCMV Antibodies Recognize the NAG-2 Receptor on Dermal Fibroblasts and Do Not Affect Cell Proliferation
(A and B) FACS analysis of the binding of antibodies to normal human fibroblasts: antibodies affinity purified against the irrelevant peptide from sera of SSc patients (A) and antibodies affinity purified against the UL94 peptide from the same sera (B). Percentage of positive cells is 7% (A) and 83 % (B), respectively.
(C) Anti-peptide antibodies react with the NAG-2 molecule. Lysates from human dermal fibroblasts were immunoprecipitated with a rabbit affinity purified anti-UL94 peptide antibody crosslinked to sepharose. Immunoprecipitates were resolved in SDS-PAGE and transferred to nitrocellullose. Blots were incubated with pre-immune rabbit antiserum (lane 1), rabbit anti-UL94 peptide antibody (lane 2), affinity purified antibodies directed against the irrelevant control peptide isolated from patients with SSc (lane 3), and anti-UL94 peptide antibodies purified from the sera of the same patients (lane 4).
(D) Binding of affinity purified antibodies against the UL94 peptide to fibroblasts is specifically inhibited by the UL94 peptide (diamonds) but not by an irrelevant control peptide (circles). Data represent percent of inhibition; inhibitor concentration (horizontal axis) is in micrograms per milliliter.
(E) Fibroblasts incubated with anti-hCMV antibodies normally proliferate. Shown are proliferation levels for fibroblasts incubated for various intervals of time with medium alone (light grey bars), with affinity purified antibodies directed against the irrelevant control peptide (dark grey bars), or with affinity purified antibodies against the UL94 peptide (black bars). OD, optical density at 570 nm.
Gene Expression Profile in HUVECs Treated with Anti-hCMV Peptide Antibodies
Since we have already shown that anti-UL94 peptide antibodies promote endothelial cell apoptosis following engagement of the NAG-2 molecule [11], we decided to analyze the gene expression profiles induced in endothelial cells by the anti-hCMV antibodies in order to identify clusters of genes known to be involved in the pathogenesis of vascular damage in SSc.
For this purpose normal endothelial cells were incubated with either anti-UL94 peptide antibodies affinity purified from the sera of patients with SSc or with control antibodies affinity purified against an irrelevant peptide from the sera of the same individuals. The gene expression profiles were studied at two different time points: after 4 and 8 h of stimulation. As stated in Methods, we considered only those genes expressed more than 2-fold above control at minimally one time point. Using these criteria, anti-hCMV antibodies were found to upregulate 1,645 transcripts (Dataset S1) including genes encoding adhesion molecules, chemokines, CSFs, growth factors, and molecules involved in apoptosis.
Figure 2 shows an overview of some genes within the above mentioned clusters. A more detailed representation of the same genes is presented in compiled form in Table 2, which includes GeneBank accession numbers and F.C. of expression of the genes. Among the genes encoding adhesion molecules, the highest increase in expression was observed for E-selectin, VCAM-1, and ICAM-1 coding genes (F.C. 68.5, 26.5, and 18.8, respectively, at 4 h of stimulation) (Table 2). High circulating levels of these adhesion molecules have been found in scleroderma [20].
Figure 2 Profile of Selected Genes Modulated in Response to Anti-hCMV Antibodies
Figure shows F.C. in gene expression in endothelial cells (A) and fibroblasts (B) at 4 and 8 h after incubation with affinity purified anti-UL94 antibodies. Genes are grouped into functional categories based on our current understanding of their function. Genes shown have at least a 2-fold change in gene expression.
Table 2 Gene Expression in HUVECs after 4 and 8 h of Antibody Stimulation
Another cluster of upregulated genes was represented by genes encoding chemokines (Table 2), a group of cytokines able to attract leukocytes and regulate angiogenesis, vascular proliferation, and fibrosis, functions that may contribute to the manifestations of scleroderma [21]. The expression of MCP-1 (also called CCL2) was highly increased, with a peak of 74-fold increase in expression at 4 h of stimulation. MCP-1 stimulates fibroblast proliferation and collagen production [22]. Several other chemokine genes were upregulated in endothelial cells treated with anti-hCMV peptide antibodies, such as those encoding MCP-3 (or CCL7), Macrophage inflammatory protein 1 alpha (MIP-1 alpha [or CCL3]), Growth regulated oncogene 1 (GRO 1 [or CXCL1]), Stromal cell-derived factor 1 (SDF-1), and Fractalkine.
Consistent with endothelial damage and activation, an increased expression of the genes encoding von Willebrand factor (vWF), Tissue factor (F3), Ephrin A1, and ET-1 was present in stimulated endothelial cells. Although the increase in expression of these genes is not specific for SSc, vWF and F3 molecules are associated with the acquisition of a procoagulant phenotype typical of activated endothelial cells and Ephrin A1 and ET-1 molecules are associated with diseases characterized by endothelial dysfunction [23]. Of note is the upregulation of the gene encoding Angiotensin II receptor type 1, which plays an important role in ischemia-induced angiogenesis [24].
Moreover, we found an increased expression of several genes encoding growth factors (Table 2), including CSFs, Fibroblast growth factor 2 (FGF-2), Platelet-derived growth factor (PDGF), and Insulin-like growth factor 2 (IGF-2). These molecules have also been implicated in the pathogenesis of SSc; for example, PDGF seems to be associated with fibrotic damage in SSc [25].
Finally, genes encoding molecules involved in the apoptotic process, such as TRAMP (DR3), GRIM19, Caspase 1, and DNAse 1 and 2, were upregulated in accordance with the observation that endothelial cells undergo cell death following incubation with anti-hCMV antibodies [11].
Taken together, these results show that a subset of anti-hCMV antibodies modulate the expression of a series of gene clusters encoding molecules playing a fundamental role in the vascular damage typical of SSc.
Gene Expression Profile in Human Dermal Fibroblasts Treated with Anti-hCMV Peptide Antibodies
Human dermal fibroblasts were incubated with either anti-UL94 peptide antibodies affinity purified from the sera of patients with SSc or with control antibodies purified against an irrelevant peptide from the sera of the same individuals. The gene expression profiles were studied at the same time points as for endothelial cells, and we considered for further analysis only those genes whose expression increased more than 2-fold above control at at least one time point.
The engagement of the NAG-2 receptor by anti-hCMV antibodies upregulated 989 transcripts (Dataset S2), including genes encoding molecules involved in extracellular matrix deposition, growth factors, chemokines, and cytokines. Figure 2A shows an overview of a panel of genes within the above mentioned clusters. A more detailed representation of the same genes is presented in compiled form in Table 3. An excessive deposition of collagen and extracellular matrix is typical of scleroderma fibroblasts and leads to fibrosis of the skin and internal organs [13]. When compared to control cells, treated fibroblasts showed increased expression of genes encoding different types of collagens such as Collagen type I, type XVI, and type XI (Table 3). Similarly several other genes involved in extracellular matrix deposition were upregulated including those encoding Fibronectin, Emilin 1, Dermatopontin, Biglycan, Cartilage oligomeric matrix protein, and Tenascin C with F.C. in expression ranging from 3.53 to 6.82. Increased levels of the above mentioned proteins have been associated with SSc [26]. Also, the gene encoding Bone morphogenetic protein 1 (Procollagen C-proteinase), an enzyme important for the formation of mature collagens [27], was upregulated, possibly because of the elevated rate of collagen synthesis observed in treated fibroblasts. A very high level of induction was observed for the Hyaluronan synthase 2 gene, with a F.C. in expression of 18.12 at 4 h, which is compatible with the function of the enzyme in extracellular matrix metabolism [28].
Table 3 Genes Expression in Fibroblasts after 4 and 8 h of Antibody Stimulation
Analysis of gene expression patterns revealed that the fibroblasts used in this study also expressed some cytoskeletal genes encoding proteins typically associated with the myofibroblast phenotype (e.g., Transgelin and Elastin) [29]. Anti-hCMV antibodies also induced the expression of a number of cytoskeletal genes, which are usually expressed by highly differentiated smooth muscle cells, such as the gene for Calponin [29]. Of interest, we also observed induction of the smooth-muscle-cell-restricted signaling molecule Cysteine and glycine-rich protein 2 (CSRP2, also known as SmLIM for smooth muscle LIM-containing protein), which is expressed in differentiated vascular smooth muscle cells [30].
Overexpression of genes coding for proinflammatory and profibrotic cytokines (Table 3) has been observed: IL-6, IL-8, and IL-11, with F.C. in expression of 6.4, 5.54, and 40.78, respectively, at 4 h. Similarly, genes encoding some chemokines (see Table 2) were upregulated, including MCP-1, MCP-3, and IP10 with F.C. in expression of 12.9, 4.6, and 12.9, respectively, at 8 h of stimulation.
Another group of genes included those encoding growth factors typical of activated fibroblasts, among them TGF-beta 1 and its accessory receptor Betaglycan (TGFBR3) [31], and Connective tissue growth factor (CTGF), which represent key mediators of fibrosis in SSc [32]. Interestingly, a series of genes known to be upregulated in TGF-beta 1–treated normal fibroblasts [29,33] were found overexpressed in dermal fibroblasts exposed to anti-hCMV antibodies. This cluster of genes included those encoding the transcription factor JUNB, the Smad co-activator Runt-related transcription factor 1 (RUNX1), and the transcriptional regulator TIEG. Most importantly, we found an increased expression of the signaling molecule Smad7 (F.C. in expression of 8.45 at 4 h of stimulation), known to be overexpressed in scleroderma fibroblasts [34]. Also, the upregulation of Angiotensin II receptor type 1 is likely to significantly amplify the profibrotic actions of TGF-beta 1 [35]. Moreover genes coding for VEGF, PDGFA, and PDGF receptor B were overexpressed in treated fibroblasts.
Finally, we observed an increased expression of the gene coding for Rac protein kinase-beta (Akt), an important regulator of cell proliferation and survival, and, interestingly, this gene is overexpressed in scleroderma fibroblasts [36]. The induction of Akt along with that of two genes involved in regulating cell growth and apoptosis, IER3 [37] and PIM-1 [38], is consistent with the observation that anti-hCMV antibodies promote fibroblast survival.
Taken together, these results showed that genes involved in synthesis of extracellular matrix components and in cell survival and proliferation were upregulated in fibroblasts exposed to anti-hCMV antibodies.
Downregulated Genes in Endothelial Cells and Fibroblasts
The engagement of the NAG-2 receptor by anti-hCMV antibodies downregulated 1,389 genes in endothelial cells and 931 genes in fibroblasts (Datasets S3 and S4). We selected a few of them, based on their functional relevance.
Table 4 shows a list of repressed genes in endothelial cells. The gene encoding the anti-apoptotic molecule BCL2 [39] was highly repressed, in keeping with the observation that endothelial cells undergo apoptosis following engagement of the NAG-2 receptor. The decreased expression of the gene encoding Endothelial nitric oxide synthase (eNOS) has already been reported in endothelial cells isolated from patients with SSc, indicating an intrinsic defect in the mechanism of nitric oxide production [40]. It is worth noting the downregulation of the Endothelin type B receptor since this receptor on endothelial cells promotes vasodilatation through release of nitric oxide and prostacyclin, increases the clearance of ET-1, and inhibits Endothelin-converting enzyme expression [41].
Table 4 Downregulated Genes in HUVECs after 4 and 8 h of Antibody Stimulation
Table 5 summarizes the downregulated genes in fibroblasts. Interestingly the Death-associated protein kinase 1 (DAPK-1), a pro-apoptotic protein, was reduced in fibroblasts with a F.C. in expression of −4.47 and −7.5 at 4 and 8 h, respectively [42]. Also genes encoding matrix metalloproteinase proteins (MMP-1, MMP-3, and MMP-10) were reduced especially at 8 h after stimulation. These enzymes are responsible for matrix degradation and turnover and therefore their reduction may contribute to the pathogenesis of fibrosis [43,44]. Among the signal transduction molecules, Smad3, Smad1, and TGF-beta receptor 2 were downregulated.
Table 5 Downregulated Genes in Fibroblasts after 4 and 8 h of Antibody Stimulation
Q-PCR Validation of Microarray Results
We performed Q-PCR using the same endothelial and fibroblast total RNA as was used for the microarray experiments (Figure 3). Overall, the Q-PCR results were concordant with the array results in six of six genes tested, in terms of significant differences in expression between cells incubated with anti-UL94 antibodies and control samples. The genes subjected to validation included those encoding MCP-1, ICAM-1, E-selectin, and VCAM-1 in endothelial cells at 4 and 8 h (Figure 3A) and the genes encoding MCP-1, ICAM-1, IL-6, and IL-8 in fibroblasts at 4 h (Figure 3B).
Figure 3 Validation of Gene Array Results by Q-PCR
Genes selected for validation by Q-PCR in endothelial cells (A) and fibroblasts (B). MCP-1, ICAM-1, E-selectin, and VCAM-1 transcripts were increased by 12.89-, 25.36-, 17.59-, and 5.86-fold, respectively, in endothelial cells incubated with the antibodies for 4 h compared with control endothelial cells (A). MCP-1, ICAM-1, IL-6, and IL-8 transcripts were increased 5.7-, 3.44-, 9.54-, and 8.22-fold in fibroblasts incubated with the antibodies compared to control fibroblasts (B). The level of transcript expression is reported on the vertical axis. GAPDH was selected as endogenous gene.
GAPDH was selected as endogenous standard, and we saw no significant changes in the Q-PCR results when the data were normalized using beta-actin, another constitutively transcribed gene.
Validation of Genes Induced by Anti-hCMV Peptide Antibodies via Analysis of Soluble Molecules Released in Cell Supernatants
The analyses of gene expression profiles and Q-PCR were paralleled by the analysis of some of the corresponding soluble mediators released by endothelial cells and fibroblasts. Table 6 shows the concentration of the molecules tested in the supernatants of cells incubated with antibodies against the irrelevant peptide and with anti-UL94 peptide antibodies. Endothelial cell supernatants were evaluated until the 24th hour of incubation, at which point apoptosis is nearly complete. In the supernatant of endothelial cells an increased concentration of the two chemokines MCP-1 and MCP-3 could be detected, in accordance with the findings observed at gene level. Moreover IL-6, IL-8, and TGF-beta 1 concentrations were high at different time points. EGF, analyzed as a negative control since the expression of the coding gene was unchanged, was not increased in the supernatant. In the supernatant of fibroblasts, we detected a very high concentration of MCP-1. Also, MCP-3 was elevated at all time points, and IP10 was released at high levels at late time points. We found increased levels of the interleukins IL-6, IL-8, and IL-11, reflecting the upregulation of the corresponding genes. Both TGF-beta 1 and VEGF concentrations were increased, and CTGF was detectable by Western blot. Finally, Collagen type I was secreted at all time points, indicating that fibroblasts were functionally active and were secreting constituents of the extracellular matrix, a finding that is compatible with the phenotype of scleroderma fibroblasts. These data show that gene overexpression is paralleled by increased secretion of the corresponding molecules in the supernatant of stimulated cells.
Table 6 Soluble Mediators Released in Cell Supernatants
Analysis of Soluble Molecule Concentrations in the Sera of SSc Patients
To investigate whether our findings from the gene array analysis, Q-PCR, and biochemical analysis of the cell supernatants have a functional correspondance in vivo, we analyzed the sera of 81 patients affected by SSc. The levels of IL-6, IL-8, IL-11, ET-1, MCP-1, MCP-3, soluble ICAM-1, soluble VCAM-1, and soluble E-selectin were detected by ELISA (Table 7). The levels of IL-6, IL-8, ET-1, soluble ICAM-1, soluble VCAM-1, VEGF, and soluble E-selectin were significantly higher in patients with SSc than in sex- and age-matched control individuals. When the limited and the diffuse forms of the disease were considered, ET-1, MCP-1, soluble ICAM-1, and soluble E-selectin were significantly higher in patients with diffuse SSc, whereas soluble VCAM-1 was higher in patients with the limited form of the disease (Table 8). MCP-3 and IL-11 levels were not different in patients and controls. These data confirmed the in vitro findings in vivo for the majority of the molecules analyzed.
Table 7 Soluble Mediator Levels in the Sera of Patients and Controls
Table 8 Soluble Mediator Levels in the Sera of Patients Affected by Limited and Diffuse SSc
Functional Analysis of Gene Expression in Endothelial Cells and Fibroblasts
GO analysis was used to organize the gene expression data into their functional relationships based on molecular-level and biologic processes, and revealed several GO groups that are of key interest in SSc endothelial cell and fibroblast pathophysiology. For endothelial cells some of the most representative groups included cell adhesion, integrin-mediated signaling pathway, and apoptosis, and for fibroblasts the groups included development and cell proliferation, cell adhesion, immune response, extracellular matrix organization and biosynthesis, and muscle development (Tables S1–S4).
Discussion
In the present report we demonstrate that the purified subset of antibodies directed against the hCMV-derived protein UL94 bind dermal fibroblasts through the surface receptor NAG-2. Following this interaction, fibroblasts do not undergo apoptosis, as endothelial cells do, but proliferate normally and acquire an activated phenotype resembling the features of “scleroderma-like” fibroblasts. Therefore, this subset of anti-hCMV antibodies seems to promote not only endothelial cell apoptosis, but also fibroblast activation upon engagement of the same receptor molecule, NAG-2. To further investigate the effects induced in these two different cellular targets by the anti-UL94 antibody population, we used a gene array approach, which allows the simultaneous detection of thousands of genes in a given sample. By this approach we found that the purified anti-hCMV antibodies modulate a vast array of genes encoding molecules that play a pivotal role in the pathogenesis of SSc.
In endothelial cells some of these genes encode molecules that are induced under the influence of proinflammatory cytokines, such as Tumor necrosis factor alpha (TNF-alpha) and IL-1 [45]. Indeed, genes encoding adhesion molecules such as ICAM-1, VCAM-1, E-selectin, and P-selectin were upregulated, and the findings were confirmed by Q-PCR. These molecules show increased expression on endothelial cells of SSc skin lesions, and the corresponding soluble forms are increased in the serum of SSc patients [20,46]. Moreover, ICAM-1 seems to correlate with the severity of the disease [47] and E-selectin with the presence of pulmonary fibrosis [48]. We detected a significant increase of these adhesion molecules both in the supernatant of cultured cells and in the sera of the patients analyzed.
Significant abnormalities in chemokine expression have been found in SSc [22,49,50], and, indeed, different chemokine-encoding genes were activated in treated endothelial cells. The gene encoding MCP-1 presented the highest rate of activation among all the genes considered. This result was confirmed by the high increase in MCP-1 transcripts shown by Q-PCR. MCP-1 plays a pivotal role in the pathogenesis of SSc, and, indeed, it is expressed at high levels by inflammatory mononuclear cells and endothelial cells in the skin of patients with SSc of recent onset, suggesting that this molecule may be involved in the first stages of the disease [49]. Moreover, elevated levels of MCP-1 are present in bronchoalveolar lavage cells from SSc patients with lung involvement, and high serum levels of MCP-1 correlate with pulmonary fibrosis, suggesting a potential role of this chemokine in the pathogenesis of lung damage [49]. Also the gene encoding MCP-3 was upregulated, and, indeed, this molecule has been found to be elevated in early-stage SSc and may act as a fibrotic mediator activating extracellular matrix gene expression in addition to promoting leukocyte trafficking [50]. A similar behavior has been observed for MIP-1 alpha [49,51] and GRO alpha [52]. The chemokine SDF-1 was also upregulated; since hypoxia is a potent stimulus for SDF-1, we can hypothesize that during early vascular damage the hypoxic microenvironment may induce the release of SDF-1, as demonstrated in another autoimmune disease, rheumatoid arthritis [53]. And lastly among chemokines, Fractalkine was highly induced in endothelial cells exposed to anti-hCMV antibodies (with a fold increase in expression of 4.5 at 4 h of treatment); this chemokine has been found to be overexpressed in endothelial cells of affected skin and in the lung tissues of patients with SSc. Soluble fractalkine levels were significantly raised in sera of patients with SSc and were associated with digital ischemia and severity of pulmonary fibrosis [54].
Among the other genes found activated, of particular relevance are the genes that encode ET-1 and PDGF. ET-1 is a potent vasoconstrictor molecule known to be associated with vascular damage in cardiovascular disease [55]. ET-1 is also able to induce fibrosis by enhancing collagen synthesis and inhibiting expression of MMP-1 in skin fibroblasts. Indeed, ET-1 induces a fibrogenic phenotype in normal fibroblasts that is similar to that of lesional SSc fibroblasts [56,57]. ET-1 has been extensively studied in patients with SSc who present high circulating levels of the molecule [58]. The profibrotic cytokine PDGF plays a major role in stimulating the replication, survival, and migration of fibroblasts during the pathogenesis of fibrotic diseases. Moreover, it seems to be able to induce a significant increase in MCP-1 messenger RNA and protein in cultured fibroblasts [59]. Finally, the upregulation of apoptosis-related genes is in accordance with the reported endothelial cell apoptosis induced by anti-UL94 antibodies.
In fibroblasts, the genes upregulated are mainly involved in cell activation and proliferation and in the production of extracellular matrix, all features of scleroderma. In particular, proinflammatory and profibrotic cytokines such as IL-6 and IL-8 may act synergistically with the chemokines MCP-1 and MCP-3 in recruiting monocytes and leukocytes at the site of inflammation and acting as fibrotic mediators. The upregulation of genes encoding IL-6, IL-8, and MCP-1 have been confirmed by the high increase in the corresponding transcripts detected by Q-PCR.
MCP-1 and MCP-3 have been shown to be overexpressed in SSc fibroblasts [22,50]. The production of MCP-1 by fibroblasts may be mediated by PDGF [59], whose gene is upregulated in treated endothelial cells and fibroblasts. Moreover, the gene encoding PDGF receptor B was overexpressed in fibroblasts treated with anti-hCMV antibodies, and it is known that PDGF action is determined by the relative expression of PDGF receptors on the surface of myofibroblasts. These receptors are induced during fibrogenesis, thereby amplifying the biological responses to PDGF.
The gene encoding the profibrotic cytokine TGF-beta 1 was also overexpressed. TGF-beta 1 represents a major stimulator of collagen gene expression through two different signaling pathways, immediate and delayed [26]. The first one involves proteins known as Smads that are phosphorylated by activated TGF-beta receptors, leading to nuclear translocation and gene transactivation. The second one involves the induction of secondary matrix stimulatory proteins by TGF-beta 1, amplifying the initial signal. This activation pathway involves CTGF [26]. Both Smad- and CTGF-encoding genes were modulated by the anti-hCMV antibodies in dermal fibroblasts. TGF-beta 1 may also induce apoptosis resistance of SSc skin fibroblasts by activating Akt, a kinase of the phosphoinositide-3-kinase/Akt signaling pathway with potent anti-apoptotic effects [36]; the resistance to apoptosis of SSc fibroblasts probably contributes to the accumulation of activated fibroblasts in the skin, promoting the deposition of extracellular matrix. Moreover, Akt has been implicated in cellular transdifferentiation; since TGF-beta 1 has been implicated in the conversion of fibroblasts to myofibroblasts, Akt activation may play a role in the increased number of myofibroblasts that is a key feature of SSc [36]. Moreover, the phosphoinositide-3-kinase/Akt pathway seems to be implicated in endothelial cell induced smooth muscle cell differentiation [60] also through ET-1 signaling [61]. Finally, EGF has been recently reported to induce upregulation of TGF-beta receptor through the phosphoinositide-3-kinase/Akt signaling pathway [62]. In accordance with characteristics of SSc, we found upregulation of the genes encoding Akt and EGF receptor, and of genes typically expressed by smooth muscle cells, in the fibroblasts exposed to anti-hCMV antibodies.
Chronic uncontrolled VEGF upregulation seems responsible for the disturbed vessel morphology in the skin of patients with SSc, and the high serum VEGF levels may be an indicator of capillary damage in SSc [63,64]. We found both overexpression of the gene encoding VEGF in cultured fibroblasts and high circulating levels of VEGF in our patients. Of particular relevance also is the upregulation of Angiotensin II receptor type 1 in endothelial cells and in fibroblasts; this receptor plays a pivotal role in ischemia-induced angiogenesis and in tissue fibrosis through excessive production of extracellular matrix components [24,35].
We also tested the level of some chemokines, cytokines, growth factors, and Collagen type I in the supernatants of stimulated and unstimulated cells and found that the concentration of the molecules measured was increased in the cells incubated with anti-hCMV antibodies, confirming that gene upregulation is paralleled by the induction of protein synthesis.
Finally, we measured the serum concentrations of some cytokines, chemokines, and adhesion molecules in patients and controls in order to confirm that the genes found overexpressed in vitro following stimulation with anti-hCMV antibodies could indeed be of relevance in vivo. We found that the levels of the majority of these molecules were significantly higher in patients than in controls, with a difference between the diffuse and limited subsets of the disease for some molecules, including MCP-1 and ET-1. A number of these soluble markers have already been reported to be increased in the serum of SSc patients [20]. The normal serum concentration of some other molecules, such as MCP-3, may be related to the presence of an increased level only in the damaged tissue, e.g., in the lungs of patients with lung fibrosis.
In conclusion, our results further support the pathogenic role of antibodies against the hCMV late protein UL94 in SSc. We found these antibodies in the vast majority of Caucasian patients with SSc from northern Italy [11]; the same antibodies have been detected in both Caucasian and African American patients, and their concentrations have been associated with the severity of the disease [65]. We show here that these anti-virus antibodies are able to induce not only endothelial cell activation and apoptosis but also fibroblast activation. They would therefore act as a unifying stimulus that may explain vascular damage and fibrosis, the two hallmarks of SSc.
Supporting Information
Dataset S1 Genes Upregulated in Endothelial Cells at 4 and 8 h of Incubation with Anti-hCMV Affinity Purified Antibodies
(689 KB XLS).
Click here for additional data file.
Dataset S2 Genes Upregulated in Human Fibroblasts at 4 and 8 h of Incubation with Anti-hCMV Affinity Purified Antibodies
(380 KB XLS).
Click here for additional data file.
Dataset S3 Genes Downregulated in Endothelial Cells at 4 and 8 h of Incubation with Anti-hCMV Affinity Purified Antibodies
(525 KB XLS).
Click here for additional data file.
Dataset S4 Genes Downregulated in Human Fibroblasts at 4 and 8 h of Incubation with Anti-hCMV Affinity Purified Antibodies
(371 KB XLS).
Click here for additional data file.
Table S1 GO Listing of Upregulated Genes in HUVEC Cells
(36 KB XLS).
Click here for additional data file.
Table S2 GO Listing of the Upregulated Genes in Fibroblasts
(36 KB XLS).
Click here for additional data file.
Table S3 GO Listing of Downregulated Genes in Endothelial Cells
(29 KB XLS).
Click here for additional data file.
Table S4 GO Listing of Downregulated Genes in Fibroblasts
(39 KB XLS).
Click here for additional data file.
Patient Summary
Background
Systemic sclerosis, or scleroderma, is the name of a progressive disease that is characterized by the abnormal growth of connective tissue and by the narrowing of small blood vessels. It is triggered when the body's immune system turns against the body, causing abnormal production of collagen, which can be limited to the skin or extend to internal organs. Two types of cells that are involved in systemic sclerosis are the endothelial cells that line the blood vessels and fibroblasts, which are involved in the increased fibrosis seen in the skin in systemic sclerosis.
Why Was This Study Done?
Previously these researchers have shown that one trigger for systemic sclerosis might be antibodies to a protein from a common human virus, cytomegalovirus, which also reacts with a molecule on the surface of endothelial cells and causes them to die. The researchers wanted to look more at this possible mechanism of disease and work out exactly how these antibodies affected endothelial cells and fibroblasts.
What Did the Researchers Do and Find?
They looked first in cells cultured in the laboratory to see if the antibody that they had found previously also stuck to fibroblasts. They found that it did, and that the attachment of this antibody caused a change in expression of many genes in both fibroblasts and endothelial cells. For example, in fibroblasts there was an increased expression of several genes that code for collagen—which is increased in fibrosis.
What Do These Findings Mean?
These findings suggest that an antibody to a common virus can trigger changes in cells similar to those seen in systemic sclerosis. Although there are no immediate implications for treatment, these results may help researchers to understand more about why systemic sclerosis develops.
Where Can I Get More Information Online?
MedlinePlus has many links to pages of information on systemic sclerosis:
http://www.nlm.nih.gov/medlineplus/scleroderma.html
The Scleroderma Foundation is a nonprofit organization based in the United States that provides information on scleroderma for patients, and supports research:
http://www.scleroderma.org/
We are indebted to Prof. Marco Cassatella for helpful discussion of the manuscript. This work was supported by grants from the Regione Veneto, Ricerca Finalizzata Venezia-Italia (CL), Cariverona Foundation (RC), Italian Ministry for Scientific Research and Technology (MURST) (AP and CL), and Giannina Gaslini Institute Foundation (AP). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Citation: Lunardi C, Dolcino M, Peterlana D, Bason C, Navone R, et al. (2006) Antibodies against human cytomegalovirus in the pathogenesis of systemic sclerosis: A gene array approach. PLoS Med 3(1): e2.
Abbreviations
CSFcolony-stimulating factor
ELISAenzyme-linked immunosorbent assay
F.C.fold change
GOGene Ontology
hCMVhuman cytomegalovirus
HUVEChuman endothelial cell
Q-PCRquantitative real-time polymerase chain reaction
SScsystemic sclerosis
==== Refs
References
Seibold JR Ruddy S Harris ED Sledge CB Scleroderma Kelly's textbook of rheumatology, 6th ed 2001 Philadelphia Saunders 1211 1239
Hesselstrand R Scheja A Shen GQ Wiik A Akesson A The association of nuclear antibodies with organ involvement and survival in systemic sclerosis Rheumatology 2003 42 534 540 12649400
Sgonc R Gruschwitz MS Dietrich H Recheis H Gershwin ME Endothelial cell apoptosis is a primary event underlying skin lesions in avian and human scleroderma J Clin Invest 1996 98 785 792 8698871
Bordron A Dueymes M Levy Y Jamin C Leroy JP The binding of some human antiendothelial cell antibodies induces endothelial cell apoptosis J Clin Invest 1998 101 2029 2035 9593758
Pandey JP LeRoy EC Human cytomegalovirus and the vasculopathies of autoimmune diseases (especially scleroderma), allograft rejection, and coronary restenosis Arthritis Rheum 1998 41 10 15 9433864
Neihart M Kuchen S Distler O Bruhlmann P Michel BA Increased serum levels of antibodies against human cytomegalovirus and prevalence of autoantibodies in systemic sclerosis Arthritis Rheum 1999 42 389 392 10025936
Muryoi T Pasturi KN Kafina MJ Cram DS Harrison LC Antitopoisomerase I monoclonal antibodies from scleroderma patients and tight skin mouse interact with similar epitopes J Exp Med 1992 175 1103 1109 1372644
Larsson K Aschan J Remberger M Ringden O Winiarski J Reduced risk for extensive chronic graft-versus-host disease in patients receiving transplants with human leukocyte antigen-identical sibling donors given polymerase chain reaction-based preemptive therapy against cytomegalovirus Transplantation 2004 77 526 531 15084929
Zhang Y McCormick LL Desai SR Wu C Gilliam AC Murine sclerodermatous graft-versus-host disease, a model for human scleroderma: Cutaneous cytokines, chemokines and immune cell activation J Immunol 2002 168 3088 3098 11884483
Ruzek MC Jha S Ledbetter S Richards SM Garman RD A modified model of graft-versus-host-induced systemic sclerosis (scleroderma) exhibits all major aspects of the human disease Arthritis Rheum 2004 50 1319 1331 15077316
Lunardi C Bason C Navone R Millo E Damonte G Systemic sclerosis immunoglobulin G autoantobodies bind the human cytomegalovirus late protein UL94 and induce apoptosis in human endothelial cells Nat Med 2000 6 1183 1186 11017152
Tarrant JM Robb L van Spriel AB Wright MD Tetraspanins: Molecular organisers of the leukocyte surface Trends Immunol 2003 24 610 617 14596886
Strehlow D Horn JH Biology of the scleroderma fibroblast Curr Opin Rheumatol 1998 10 572 578 9812218
Pannu J Trojanowska M Recent advances in fibroblast signaling and biology in scleroderma Curr Opin Rheumatol 2004 16 739 745 15577613
Subcommittee for Scleroderma Criteria of the American Rheumatism Association Diagnostic and Therapeutics Criteria Committee Preliminary criteria for the classification of systemic sclerosis (scleroderma) Arthritis Rheum 1980 21 581 590
Wellings DA Atherton E Standard FMOC protocols Methods Enzymol 1997 289 44 67 9353717
Lunardi C Bason C Leandri M Navone R Lestani M Autoantibodies to inner ear and endothelial antigens in Cogan's syndrome Lancet 2002 360 915 921 12354474
Irizarry RA Hobbs B Collin F Beazer-Barclay YD Antonellis KJ Exploration, normalization, and summaries of high density oligonucleotide array probe level data Biostatistics 2003 4 249 264 12925520
Irizarry RA Bolstad BM Collin F Cope LM Hobbs B Summaries of Affymetrix GeneChip probe level data Nucleic Acids Res 2003 31 e15 12582260
Andersen GN Caidahl K Kazzam E Petersson AS Waldenstrom A Correlation between increased nitric oxide production and markers of endothelial activation in systemic sclerosis Arthritis Rheum 2000 43 1085 1093 10817563
Atamas SP White B The role of chemokines in the pathogenesis of scleroderma Curr Opin Rheumatol 2003 15 772 777 14569209
Yamamoto T Eckes B Hartmann K Krieg T Expression of monocyte chemoattractant protein-1 in the lesional skin of systemic sclerosis J Dermatol Sci 2001 26 133 139 11378330
Ohura N Yamamoto K Ichioka S Sokabe T Nakatsuka H Global analysis of shear-responsive genes in vascular endothelial cells J Atheroscler Thromb 2003 10 304 313 14718748
Sasaki K Murohara T Ikeda H Sugaya T Shimada T Evidence for the importance of angiotensin II type 1 receptor in ischemia-induced angiogenesis J Clin Invest 2002 109 603 611 11877468
Atamas SP White B Cytokine regulation of pulmonary fibrosis in scleroderma Cytokine Growth Factor Rev 2003 14 537 550 14563355
Widom RL Regulation of matrix biosynthesis and degradation in systemic sclerosis Curr Opin Rheumatol 2000 12 534 539 11092204
Lee S Solow-Cordero DE Kessler E Takahara K Greenspan DS Transforming growth factor-beta regulation of bone morphogenetic protein-1/procollagen C-proteinase and related proteins in fibrogenic cells and keratinocytes J Biol Chem 1997 272 19059 19066 9228090
Weigel PH Hascall VC Tammi M Hyaluronan synthases J Biol Chem 1997 272 13997 14000 9206724
Chambers RC Leoni P Kaminski N Laurent GJ Heller R Global expression profiling of fibroblast responses to transforming growth factor-β1 reveals the induction of inhibitor of differentiation-1 and provides evidence of smooth muscle cell phenotypic switching Am J Pathol 2003 162 533 546 12547711
Jain MK Kashiki S Hsieh CM Layne MD Yet SF Embryonic expression suggests an important role for CRP2/SmLIM in the developing cardiovascular system Circ Res 1998 83 980 985 9815145
Kawakami T Ihn H Xu W Smith E LeRoy C Increased expression of TGF-beta receptors by scleroderma fibroblasts: Evidence for contribution of autocrine TGF-beta signaling to scleroderma phenotype J Invest Dermatol 1998 110 47 51 9424086
Shi-wen X Pennington D Holmes A Leask A Bradham D Autocrine overexpression of CTGF maintains fibrosis. RDA analysis of fibrosis genes in systemic sclerosis Exp Cell Res 2000 259 213 224 10942593
Renzoni EA Abraham DJ Howat S Shi-Wen X Sestini P Gene expression profiling reveals novel TGFb targets in adult lung fibroblasts 2004 Respir Res 5 24 Available: http://respiratory-research.com/content/5/1/24 . Accessed 11 October 2005
Asano Y Ihn H Yamane K Kubo M Tamaki K Impaired Smad7-Smurf-mediated negative regulation of TGF-beta signalling in scleroderma fibroblasts J Clin Invest 2004 113 253 264 14722617
Kawaguchi Y Takagi K Hara M Fukasawa C Sugiura T Angiotensin II in the lesional skin of systemic sclerosis patients contributes to fibrosis via angiotensin II type 1 receptors Arthritis Rheum 2004 50 216 226 14730619
Jun JB Kuechle M Min J Shim SC Kim G Scleroderma fibroblasts demonstrate enhanced activation of akt (protein kinase B) in situ J Invest Dermatol 2005 124 298 303 15675946
Sasada T Takedatsu H Azuma K Koga M Maeda M Immediate early response gene X-1, a stress-inducible antiapoptotic gene, encodes cytotoxic T-lymphocyte (CTL) epitopes capable of inducing human leukocyte antigen-A33-restricted and tumor-reactive CTLs in gastric cancer patients Cancer Res 2004 64 2882 2888 15087407
Hakansson P Segal D Lassen C Gullberg U Morse HC 3rd Dentification of genes differentially regulated by the P210 BCR/ABL1 fusion oncogene using cDNA microarrays Exp Hematol 2004 32 476 482 15145216
Karsan A Yee E Harlan JM Endothelial cell death induced by tumor necrosis factor-alpha inhibited by the Bcl-2 family member, A1 J Biol Chem 1996 271 27201 27204 8910286
Romero LI Zhang DN Cooke JP Ho HK Avalos E Differential expression of nitric oxide by dermal microvascular endothelial cells from patients with scleroderma Vasc Med 2000 3 147 158
Schachna L Wigley FM Targeting mediators of vascular injury in scleroderma Curr Opin Rheumatol 2002 14 686 693 12410092
Jang CW Chen CH Chen CC Chen JY Su YH TGF-beta induce apoptosis through Smad-mediated expression of DAP-kinase Nat Cell Biol 2002 4 51 58 11740493
Sato S Hayakawa I Hasegawa M Fujimoto M Takehara K Function blocking autoantibodies against matrix metalloproteinase-1 in patients with systemic sclerosis J Invest Dermatol 2003 120 542 547 12648215
Nishijima C Hayakawa I Matsushita T Komura K Hasegawa M Autoantibody against matrix metalloproteinase-3 in patients with systemic sclerosis Clin Exp Immunol 2004 138 357 363 15498049
Holzman LB Marks RM Dixit M A novel immediate-early response gene of endothelium is induced by cytokines and encodes a secret protein Mol Cell Biol 1990 10 5830 5838 2233719
Sato S Abnormalities of adhesion molecules and chemokines in scleroderma Curr Opin Rhematol 1999 11 503 507
Denton CP Bickerstaff MC Shiwen X Carulli MT Haskard DO Serial circulating adhesion molecule levels reflect disease severity in systemic sclerosis Br J Rheumatol 1995 34 1048 1054 8542206
Ihn H Sato S Fujimoto K Takehara K Tamaki K Increased serum levels of soluble vascular adhesion molecules-1 and E-selectin in patients with systemic sclerosis Br J Rheumatol 1998 37 1188 1192 9851267
Hasegawa M Sato S Takehara K Augmented production of chemokines (monocyte chemotactic protein-1 (MCP-1), macrophage inflammatory protein-1alpha (MIP-1alpha) and MIP-1beta) in patients with systemic sclerosis: MCP-1 and MIP-1alpha may be involved in the development of pulmonary fibrosis Clin Exp Immunol 1999 117 159 165 10403930
Ong VH Evans LA Shiwen X Fisher IB Rajkumar V Monocyte chemoattractant protein 3 as a mediator of fibrosis: Overexpression in systemic sclerosis and the type 1 tight-skin mouse Arthritis Rheum 2003 48 1979 1991 12847692
Bolster MB Ludwicka A Sutherland SE Strange C Silver RM Cytokine concentrations in bronchoalveolar lavage fluid of patients with systemic sclerosis Arthritis Rheum 1997 40 743 751 9125259
Furuse S Fujii H Kaburagi Y Fujimoto M Hasegawa M Serum concentrations of CXC chemokines interleukin 8 and growth regulated oncogen alpha are elevated in patients with SSc J Rheumatol 2003 30 1524 1528 12858452
Hitchon C Wong K Ma G Reed J Lyttle D Hypoxia-induced production of stromal cell-derived factor 1 (CXCL12) and vascular endothelial growth factor by synovial fibroblasts Arthritis Rheum 2002 46 2587 2597 12384916
Hasegawa M Sato S Echigo T Hamaguchi Y Yasui M Up regulated expression of fractalkine/CX3CL1 and CX3CR1 in patients with systemic sclerosis Ann Rheum Dis 2005 64 21 28 15608300
Mohacsi A Magyar J Tamas B Nanasi PP Effects of endothelins on cardiac and vascular cells: New therapeutic target for the future? Curr Vasc Pharmacol 2004 2 53 63 15320833
Kahaleh MB Endothelin, an endothelial dependent vasoconstrictor in scleroderma. Enhanced production and prophibrotic action Arthritis Rheum 1991 34 978 983 1859492
Shi-Wen X Denton CP Dashwood MR Holmes AM Bou-Gharios G Fibroblast matrix gene expression and connective tissue remodeling: Role of endothelin-1 J Invest Dermatol 2001 116 417 425 11231316
Vancheeswaran R Magoulas T Efrat G Wheeler-Jones C Olsen I Circulating endothelin-1 levels in systemic sclerosis subsets—A marker of fibrosis or vascular dysfunction? J Rheumatol 1994 21 1838 1844 7837147
Distler O Pap T Kowal-Bielecka O Meyringer R Guiducci S Overexpression of monocyte chemoattractant protein 1 in systemic sclerosis: Role of platelet-derived growth factor and effects on monocyte chemotaxis and collagen synthesis Arthritis Rheum 2001 44 2665 2678 11710722
Brown DJ Rzucidlo EM Merenick BL Wagner RJ Martin KA Endothelial cell activation of the smooth muscle cell phosphoinositide 3-kinase/Akt pathway promotes differentiation J Vasc Surg 2005 41 509 516 15838487
Shi-Wen Y Chen Y Denton CP Easwood M Renzoni EA Endothelin-1 promotes myofibroblast induction through the ETA receptor via rac/phosphoinositide 3-kinase/Akt-dependent pathway and is essential for the enhanced contractile phenotype of fibrotic fibroblasts Mol Biol Cell 2004 15 2707 2719 15047866
Yamane K Ihn H Tamaki K Epidermal growth factor up-regulates expression of transforming growth factor beta receptor type II in human dermal fibroblasts by phosphoinositide 3-kinase/Akt signalling pathway: Resistance to epidermal growth factor stimulation in scleroderma fibroblasts Arthritis Rheum 2003 48 1652 1666 12794834
Distler O Distler JH Scheid A Acker T Hirth A Uncontrolled expression of vascular endothelial growth factor and its receptors leads to insufficient skin angiogenesis in patients with systemic sclerosis Circ Res 2004 95 109 116 15178641
Choi JJ Min DJ Cho ML Elevated vascular endothelial growth factor in systemic sclerosis J Rheumatol 2003 30 1529 1533 12858453
Namboodiri AM Rocca KM Pandey J IgG antibodies to human cytomegalovirus late protein U94 in patients with systemic sclerosis Autoimmunity 2004 37 241 244 15497459
|
16318412
|
PMC1298939
|
CC BY
|
2021-01-05 10:40:33
|
no
|
PLoS Med. 2006 Jan 6; 3(1):e2
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030002
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1631841310.1371/journal.pmed.0030004Research ArticleInfectious DiseasesEpidemiology/Public HealthHealth EconomicsHealth PolicyHIV/AIDSSexual HealthHIV Infection/AIDSMedicine in Developing CountriesHealth EconomicsChronic Disease ManagementResource allocation and rationingCost-Effectiveness of Highly Active Antiretroviral Therapy in South Africa Cost-Effectiveness of HAART in RSABadri Motasim
1
*Maartens Gary
2
Mandalia Sundhiya
3
4
Bekker Linda-Gail
1
Penrod John R
5
Platt Robert W
6
Wood Robin
1
Beck Eduard J
3
6
1Department of Medicine, Desmond Tutu HIV Centre, University of Cape Town, Cape Town, South Africa2Department of Clinical Pharmacology, University of Cape Town, Cape Town, South Africa3NPMS-HHC Coordination and Analytic Centre, Chelsea and Westminster Hospital, London, United Kingdom4Department of Medicine, Imperial College, Chelsea and Westminster Hospital, London, United Kingdom5Division of Clinical Epidemiology, McGill University Health Centre, Montreal, Quebec, Canada6Department of Epidemiology and Biostatistics, McGill University, Montreal, Quebec, CanadaCarr Andrew Academic EditorSt. Vincent's HospitalAustralia*To whom correspondence should be addressed. E-mail: [email protected]
Competing Interests: Study was funded by Secure the Future, Bristol-Myers Squibb, through an unrestricted educational grant.
Author Contributions: MB, GM, SM, LB, JRP, RWP, RW, and EJB contributed to the conceptualization of the study. MB, GM, SM, RW, and EJB designed the study. GM, LB, and RW enrolled patients. MB and SM performed the statistical analysis. MB wrote the first draft of the paper. All authors contributed to the writing of the final version of the manuscript.
1 2006 6 12 2005 3 1 e417 5 2005 27 9 2005 Copyright: © 2006 Badri et al.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
HAART: A Cost-Effective Option for South Africa
Background
Little information exists on the impact of highly active antiretroviral therapy (HAART) on health-care provision in South Africa despite increasing scale-up of access to HAART and gradual reduction in HAART prices.
Methods and Findings
Use and cost of services for 265 HIV-infected adults without AIDS (World Health Organization [WHO] stage 1, 2, or 3) and 27 with AIDS (WHO stage 4) receiving HAART between 1995 and 2000 in Cape Town were compared with HIV-infected controls matched for baseline WHO stage, CD4 count, age, and socioeconomic status, who did not receive antiretroviral therapy (ART; No-ART group). Costs of service provision (January 2004 prices, US$1 = 7.6 Rand) included local unit costs, and two scenarios for HAART prices for WHO recommended first-line regimens: scenario 1 used current South African public-sector ART drug prices of $730 per patient-year (PPY), whereas scenario 2 was based on the anticipated public-sector price for locally manufactured drug of $181 PPY. All analyses are presented in terms of patients without AIDS and patients with AIDS.
For patients without AIDS, the mean number of inpatient days PPY was 1.08 (95% confidence interval [CI]: 0.97–1.19) for the HAART group versus 3.73 (95% CI: 3.55–3.97) for the No-ART group, and 8.71 (95% CI: 8.40–9.03) versus 4.35 (95% CI: 4.12–5.61), respectively, for mean number of outpatient visits PPY. Average service provision PPY was $950 for the No-ART group versus $1,342 and $793 PPY for the HAART group for scenario 1 and 2, respectively, whereas the incremental cost per life-year gained (LYG) was $1,622 for scenario 1 and $675 for scenario 2. For patients with AIDS, mean inpatients days PPY was 2.04 (95% CI: 1.63–2.52) for the HAART versus 15.36 (95% CI: 13.97–16.85) for the No-ART group. Mean outpatient visits PPY was 7.62 (95% CI: 6.81–8.49) compared with 6.60 (95% CI: 5.69–7.62) respectively. Average service provision PPY was $3,520 for the No-ART group versus $1,513 and $964 for the HAART group for scenario 1 and 2, respectively, whereas the incremental cost per LYG was cost saving for both scenarios. In a sensitivity analysis based on the lower (25%) and upper (75%) interquartile range survival percentiles, the incremental cost per LYG ranged from $1,557 to $1,772 for the group without AIDS and from cost saving to $111 for patients with AIDS.
Conclusion
HAART is a cost-effective intervention in South Africa, and cost saving when HAART prices are further reduced. Our estimates, however, were based on direct costs, and as such the actual cost saving might have been underestimated if indirect costs were also included.
Healthcare costs for HIV-infected South African adults on HAART compared with costs for HIV-infected controls not on HAART. Authors conclude HAART is cost effective.
==== Body
Introduction
South Africa is experiencing an HIV epidemic with enormous social and economic consequences. Recent estimates suggest that between 4.5 and 6.2 million of the 43 million South Africans are infected with HIV-1 [1]. There were 370,000 AIDS deaths during 2003 [1], and the cumulative projected AIDS mortality for 2010 is 4–7 million in absence of a highly active antiretroviral therapy (HAART) programme [2]. The largest impact of HIV on the public health sector lies in the hospital sector [3]. In the year 2000, HIV-related admissions amounted to 24% of all public hospital admissions [4] and 12.5% of the total public health budget [5]. Cost of inpatient and ambulatory health care of both private and public health-care sectors is expected to rise rapidly [5].
The cost-effectiveness of HAART, in terms of reducing HIV-related morbidity and mortality, has been documented in industrialized countries [6–12]. The introduction of combination HAART into routine clinical care in these countries has been associated with a shift from inpatient to outpatient-based hospital care [11–17]. Until recently the prevailing assumption was that the public sector of the South African health-care system was unable to afford the introduction of antiretroviral therapy (ART) in routine clinical care. However, the government of South Africa recently announced its commitment towards creating the necessary conditions for introducing ART into the public health sector [18]. In addition, the price of HAART for resource-poor countries decreased markedly since the year 2000 [19,20]. The South African Department of Health has recently awarded contracts for the supply of ART drugs to public health facilities countrywide to international pharmaceutical companies [21]. This tender is expected to reduce HAART price to $181 per patient-year (PPY).
The aim of this study was to compare use and cost of HIV-1–related service provision between patients receiving HAART and a comparison group not receiving ART, and assess the cost effectiveness of HAART.
Methods
Study Population
This study was based on the Cape Town AIDS Cohort (
CTAC); a prospective cohort study which has been described previously [22,23]. In brief, patients of this cohort were accrued from the HIV clinics affiliated to the University of Cape Town, who were referred from a wide range of primary HIV health-care providers. During the study period 1st January 1995 to 31st December 2000, HAART was not available in the publicly funded South African health-care sector. All patients in this study accessed HAART through the participation in the international HAART multicentre phase III clinical trials, as approved by the Research Ethics Committee of the University of Cape Town.
For the purpose of this study, all patients who participated in the clinical trials and received at least three ART drugs—a non-nucleoside reverse transcriptase inhibitor or protease inhibitor together with two nucleoside analogues or three nucleoside analogues—were included as the treated arm of the study (HAART group). Patients were excluded from the clinical trials if they were active injecting-drug users, were diagnosed with an acute opportunistic infection at the time of recruitment, were reported to have significant laboratory abnormalities, or if they were treated with immune-modulating or systemic chemotherapeutic agents. Lactating or pregnant women were also excluded. The trial visit schedule was usually at weeks 2, 4, and 8 and then every two to three months thereafter.
Patients who did not participate in these clinical trials and never had access to ART throughout the study period (No-ART group) but received other HIV-related care were the sample from which a “comparator” group was identified for the HAART group.
At each clinic visit, all patients were routinely examined for HIV related manifestations and staged using the World Health Organization (WHO) clinical HIV staging system [24]. HIV-1 infection was diagnosed by enzyme-linked immunosorbent assay (ELISA) tests and confirmed by Western blot or a second enzyme-linked immunosorbent assay test. Viral load (which was available only for the HAART group) was determined by reverse transcriptase-polymerase chain reaction (Amplicor; Roche Molecular Systems, Branchburg, New Jersey, United States) and CD4+ count, measured by flow cytometry (Beckman Coulter, Miami, Florida, United States).
Analysis
This analysis calculated the use and cost of HIV service provision and compared the clinical outcome, in terms of disease progression or life year gained (LYG) by clinical stage of HIV infection, between patients receiving HAART and a matched comparison group who did not receive ART (No-ART group). Patients were classified as either being non-AIDS (WHO stages 1, 2, or 3) or AIDS (WHO stage 4) patients.
Several strategies were employed to ensure that the two groups studied were clinically, immunologically, and socioeconomically similar and matched for the same variables used to recruit the HAART group into the clinical trials. Logistic regression models were fitted to identify factors associated with receiving HAART in this cohort using SAS GENMODE procedure with logit link function and binomial error distribution [25]. HAART patients were individually matched with randomly selected No-ART patients on the basis of variables independently associated with the likelihood of receiving HAART. The socioeconomic status of each patient was classified into “low” or “high”, using a composite index developed by the Cape Metropolitan Council [26]. A subgroup logistic regression analysis was performed for the HAART group, to examine whether the likelihood of hospitalisation differed by HAART class.
To examine for residual confounding, the matched case–control data were analysed using a conditional logistic regression model, stratified by matching variables. The model was fitted using the SAS PHREG procedure with discrete logistic model. All data analyses were performed in SAS version 8.02. χ2 was used to compare categorical variables, and the non-parametric median test was used to compare continuous non-normally distributed variables. All p-values quoted are two sided, with a p-value < 0.05 considered as significant.
Use and Cost of Services
Information on inpatient and outpatient care was obtained from the computerized hospital information systems supplemented by case notes. The mean number of inpatient days and outpatient visits PPY were calculated for the non-AIDS (WHO stage 1, 2, and 3) and AIDS (WHO stage 4) WHO clinical stages for both HAART and No-ART groups. A patient-year was defined as 365.25 days of follow up and methods used for calculating the mean use of services were similar to those used in other studies [12,16,17,27]. The denominator consisted of the total duration of follow up for all patients seen during the study period and numerators were calculated by summing the use of each service. Mean and 95% confidence intervals (CI) of inpatient and outpatient service use PPY by WHO stage were calculated for the two groups using the binomial distribution, and were compared between the two groups by calculating the odds ratio (OR) of the use of inpatient and outpatient services, using the No-ART group as a reference group.
The costs of hospital HIV service provision were calculated from a public health-care system perspective [28–30]. Unit costs were obtained from a detailedcosting study of HIV inpatient and outpatient care conducted in the year 2000 [31], and were adjusted for inflation to financial year 2004 prices using the South African Consumer Price Index [32]. Prices were converted from South African Rand to US dollars using the average exchange rate for 2004 (US dollars = 7.6 Rand) [33]. The unit cost was $215 for an inpatient day and $33 for an outpatient visit and included costs for tests including CD4 counts, procedures, and non-ART drug costs. The non-ART drugs included all drugs other than ART dispensed to the patients during the course of care, including treatment and prophylaxis for opportunistic infections. Mean inpatient days and outpatient visits PPY were multiplied by their respective unit costs to estimate the PPY cost of service provision.
ART prices used in this study are those currently available to the public health-care sector (Ministry of Health, Provincial Administration of the Western Cape). HAART drug-price scenarios presented were (1) present public sector prices, which amounted to $730 per annum, and (2) anticipated public sector price for locally manufactured drugs, which amounted to $181 per annum, for the WHO-recommended regimen for resource-limited settings [34].
To estimate the total cost of service provision PPY for HAART patients for the two scenarios, average ART drug costs PPY were added to the average inpatient and outpatient PPY costs. In sensitivity analysis, minimum and maximum ART drug PPY costs for the two scenarios were also added to the lower and upper limit of the 95% CI: inpatient and outpatient PPY cost of care to provide a range of costs. Viral load was not measured for the No-ART group because it was not available in publicly funded institutions during the study period and, therefore, PPY cost of viral-load investigation of $79 (D. Roditti, personal communication) was only added to the annual cost of service provision for the HAART group.
Cost of LYG by WHO Stage of HIV Infection
Progression times were calculated from date of entry into non-AIDS (WHO stage 1, 2, or 3) to date of progression to AIDS (WHO stage 4) or death, and from initial diagnosis of AIDS (WHO stage 4) to death for AIDS patients. Patients not known to have progressed during follow-up were censored at either the most recent visit to the clinic or when lost to follow-up. Median progression times were estimated using the product-limit Kaplan-Meier survival method, and these were compared for the HAART and No-ART groups using log-rank test. Due to the small number of individuals who progressed during the follow-up period, median and interquartile ranges (IQR) for time to progression to AIDS or death were extrapolated from the product-limit time to failure estimates using the maximum likelihood least squares method. The progression-free times for non-AIDS and AIDS patients for each group were multiplied by the average PPY cost of service provision, and the additional life years gained of non-AIDS and AIDS groups was calculated as the incremental cost per LYG, based on the difference in the estimated median progression times of the two groups [27].
Because discounting health benefits remains controversial [35], only non-discounted estimates are presented. However, given the relatively short time in each WHO stage, it is unlikely that an analysis with a non-zero discount rate would yield qualitatively different results than those presented here.
Sensitivity Analysis
Robustness of results was assessed in a sensitivity analysis; accounting for variances associated with treatment effects and total cost of service provision. IQRs between the lower (25%) and upper (75%) progression-free times percentiles of the non-AIDS and AIDS patients were multiplied by the average and 95% CI of the cost of service provision, and the incremental cost per LYG was calculated.
Results
Study Sample
Of the 1,630 patients in the cohort, 292 patients (265 non-AIDS and 27 with AIDS) received HAART through participation in the clinical trials. The rest of the patients (n = 1,328; 1,093 non-AIDS and 235 with AIDS) did not have access to ART during the study period and comprised the population from which the No-ART comparator group for the 292 patients who received HAART was identified. Baseline CD4 count, WHO stage, age, and socioeconomic status were independently associated with the likelihood of receiving HAART (Table 1), but gender was not, and therefore this variable was not considered in further analyses. Matching was therefore based on WHO stage, CD4 count (<200, 200–350, and >350 cells/μl), age (less than the median age or equal to the median age or greater of the non-AIDS and AIDS groups respectively) and socioeconomic status (low or high socioeconomic status).
Table 1 Univariate and Multivariate Logistic Regression Analyses of Factors Associated with the Likelihood of Receiving HAART
HAART drug classes were not independently associated with increased risk of hospitalisation (Table 2) and were therefore analysed as one category. The characteristics of the final study population of the 292 patients who received HAART and the 292 matched No-ART patients are described in Table 3.
Table 2 Logistic Regression Analysis of Factors Associated with Hospitalisation among the Treated Group
Table 3 Baseline Demographic and Clinical Characteristics of the Matched Non-AIDS and AIDS Groups
The Non-AIDS Population (WHO Stage 1, 2, or 3)
The matched non-AIDS group included 265 patients both in the HAART and No-ART group. Approximately one-third of the patients in the two groups had a baseline CD4 count <200 cell/μl and (49.4%) were of low socioeconomic status. Median age at inclusion into study did not differ in the two groups; 32 y, [IQR: 28–39 y] in the HAART group versus 32 y [IQR: 28–40 y] in the No-ART group (median test p = 0.48). Although not matched for, gender distribution did not differ statistically in the two groups (χ2 = 0.07, p = 0.79; Table 3). Median progression time was significantly longer in the HAART group compared with the No-ART group at 4.1 and 3.0 y respectively (log-rank test χ2 = 36.6, p < 0.001; Figure 1).
Figure 1 Progression of HIV-Infected Individuals from Non-AIDS Stages (WHO Stage 1, 2, or 3) for Patients on HAART and Not on ART
The solid line indicates patients on HAART, and the dotted line indicates patients not on ART.
Use and Cost of Services and Cost per LYG
Patients on HAART had 1.08 (95% CI: 0.97–1.19) mean inpatient days, significantly fewer than the 3.75 d (95% CI: 3.55–3.97) of the No-ART group; χ2 = 147, OR = 0.29, 95% CI: 0.23–0.36, p < 0.001; but had significantly more outpatient visits of 8.71 (95% CI: 8.40–9.03) compared with 4.35 (95% CI: 4.12–5.61); χ2 = 145, OR = 2.00, 95% CI: 1.78–2.25, p < 0.0001 (Table 4). The average PPY inpatient cost in the HAART group was significantly less than that for the No-ART group, while the average costs of outpatient visits PPY for the No-ART group were less than those for the HAART group (Table 4).
Table 4 Mean Number of Inpatient Days, Outpatient Visits, and Associated Costa PPY
Based on the two HAART price scenarios, the average cost of service provision PPY for the HAART group ranged from a minimum of $760 to $1,377 PPY, with scenario 2 having the lowest service provision cost (Table 4). The incremental cost per LYG for median progression time was $1,622 (95% CI: 1,607–1,627) for scenario 1 and $675 (95% CI: 659–679) for scenario 2 (Table 5). When a sensitivity analysis was performed based on the IQR of the progression times, the incremental cost per LYG varied between $1,578 (95% CI: 1,557–1,581) and $1,759 (95% CI: 1,748–1,772) for the 25th and 75th percentiles respectively (Table 5).
Table 5 Incremental Cost-Effectiveness Ratio (US$) for Current ART Rollout Prices (US$730 Per Annum—Scenario 1) and Anticipated Tender Prices (US$181 Per Annum—Scenario 2), Comparing HAART and No-ART Groups for Non-AIDS and AIDS Groups at 25th, 50th (Median), and 75th Progression-Free Times Percentiles
The AIDS Population (WHO Stage 4)
The AIDS population included 27 patients in each group. The majority of patients in the two groups presented with a CD4 count <200 cell/μl (77%), and 40.74% were of low socioeconomic status. Median age did not differ in the two groups; 35 y (IQR: 32–41) in the HAART group versus 37 y (IQR: 33–50) in the No-ART group (median test p = 0.27). Gender distribution, with 63% and 70.4% males in the HAART and No-ART groups respectively, was not significantly different in the two groups (χ2 = 0.33, p = 0.56) (see Table 3). Median progression time was significantly longer in the HAART group compared with the No-ART group; at 3.1 and 1.4 y respectively (log-rank χ2 = 5.28, p = 0.02; Figure 2).
Figure 2 Progression of HIV-Infected Individuals from WHO Stage 4 for Patients on HAART and Not on ART
The solid line indicates patients on HAART, and the dotted line indicates patients not on ART.
Use and Cost of Services and Cost per LYG
Patients on HAART had significantly fewer mean PPY inpatient days at 2.04 d (95% CI: 1.63–2.52) compared with 15.36 d (95% CI: 13.97–16.85) for the No-ART group (χ2 = 1,019, OR = 0.13, 95 CI: 0.11–0.15, p < 0.0001). Mean outpatient visits PPY in the two groups did differ significantly; at 7.62 (95% CI: 6.81–8.49) for the HAART group compared with 6.60 (95% CI: 5.69–7.62) for the No-ART group; χ2 = 7.3, OR = 1.15, 95% CI: 1.04–1.28, p = 0.007, though not as substantially as for the non-AIDS group (see Table 4). The average inpatient cost PPY in the HAART group was significantly less than that for the No-ART group, but the average costs of outpatient visits PPY in the groups were not significantly (see Table 4).
Based on the two HAART price scenarios, the average cost of service provision PPY for the HAART group ranged from a minimum of $850 to $1,645 PPY, with the lowest care cost observed for scenario 2 (see Table 4). For patients diagnosed with AIDS, the incremental cost per LYG for the median progression time was cost saving for both HAART price scenarios (Table 5). When a sensitivity analysis was performed based on the IQR of the progression times, the incremental cost per LYG varied between $71 (95% CI: 43–111) and cost saving for the 25th and 75th progression-free time percentiles respectively (Table 5).
Discussion
This study, employing methods used in similar studies from industrialized countries [27], provides a unique contemporaneous comparison of the use, cost, and outcome of hospital service provision for a group of HIV-infected patients in Cape Town receiving HAART compared with an immunologically, clinically, and socioeconomically similar group of patients who did not receive ART. Use of HAART was associated with decreased disease progression, AIDS, and death. The HAART group used fewer inpatient services than the No-ART group, and the magnitude of these changes did not differ by HAART regimens used in this study. The reduction in use of inpatient services, which has been observed in similar studies in industrialized countries [10–15], was most likely due to a reduction in morbidity and mortality [6,12]. The use of services increased for both groups with increased severity of HIV infection, resulting in an increased cost of service provision. The increased use of inpatient services for patients with AIDS is most likely related to AIDS-related events or their terminal phase of their illness [36–41]. In Zimbabwe, medical insurance claims of privately insured HIV-infected patients in the last few months of their lives were 700% higher than that of uninfected patients in the same age group [42].
To date, very few cost-effectiveness studies have been performed on HAART in a South African setting [43]. The incremental cost per LYG ranged from being cost saving to $1,772. The cut-off point for what constitutes a cost-effective intervention per outcome measure varies from society to another. For instance, the cost-effective cut-off point in the United States is currently considered to be $50,000 per outcome measure and £30,000 in the United Kingdom [30]. To date such a consensus on what would constitute a realistic threshold for South Africa has not yet emerged, but a cut-off of twice the per capita gross domestic product (GDP) has been suggested as a reasonable cut-off point for developing countries [44]. For the year 2004, the per capita gross domestic product in South Africa was $3,480, and therefore this threshold would amount to $6,960 [45]. The cost per LYG of two HAART cost scenarios for the non-AIDS and AIDS patients showed that introducing HAART in this hospital setting would be a very cost-effective intervention. However, it is clear that the cost-effectiveness ratios were very sensitive to the price of HAART. If prices of the awarded tender could be achieved, the introduction of HAART will be a very cost-effective intervention in Cape Town and probably in similar settings in sub-Saharan Africa, because HIV accounts for between 40% and 70% of the public sector inpatient service provision in the region [3,36–40].
Concern has been expressed that increased access to HAART in sub-Saharan Africa will result in the widespread viral resistance due to poor adherence [46]. Studies performed in a number of sub-Saharan African countries, however, have shown that the proportion of individuals maintaining viral suppression is comparable to that reported from developed countries [47–49].
This study did have a number of limitations. Because HAART was not used in routine clinical practice, we had to compare a group of patients enrolled in clinical trials with a control group that was not part of the trials. Individuals who take part in clinical trials have to fulfil certain entry criteria, as well as to conform to well-defined protocols and scheduled attendances. It is therefore difficult to exclude the possibility that a selection bias might have resulted from the inclusion/exclusion criteria of these clinical trials. However, the No-ART control group was selected on the basis of clinical, socioeconomic, and immunologic characteristics similar to those individuals recruited into the HAART trials conducted in this study. The frequency of inpatient and outpatient services utilization of the HAART and No-ART patients in this study is similar to that reported by UK and Canadian observational studies [12,27]. However, in this study, the sample is relatively small for the AIDS group.
This study was focused on hospital services provided at the level of a teaching hospital. Therefore the costs incurred through primary, community, or secondary hospital care were not included, but this reflected the configuration of services available to the majority of HIV-infected people in Cape Town at the time of the study. Similarly, the costs included were direct costs only and did not incorporate the indirect or intangible costs, such as loss of productivity or quality of life associated with this illness, because currently no such data exist in South Africa. Some studies from the United Kingdom have demonstrated that from a public sector perspective, indirect costs can comprise between 58% and 124% of direct treatment costs for HAART or between 45% and 102% from a societal perspective [30]. If these costs were all included, it is likely that the cost-effectiveness ratio would even be more favourable. Our estimates did not incorporate the costs of providing the infrastructure required to support appropriate HAART provision in rollout programmes. However the rollout programmes were designed to start from settings where infrastructure currently exists, which would predominantly be urban. Recent reports estimated that if the public sector included HAART as part of a package of HIV treatment and care in the year 2003, the cost would be 1.2% of the South African GNP, which is unlikely to push health-care expenditure beyond prudent levels [50,51].
The recent commitment towards scale-up of HAART in South Africa as part of HIV treatment and care has been an important and positive development. The urgent need to introduce HAART as part of routine HIV treatment and care was recently re-iterated in a World Bank report, which indicated that if this is not done soon, failure to do so would have devastating effects on this and future generations of South Africans [52].
Although the primary rationale for wider access to HAART is humanitarian, a national HAART programme targeting patients with symptomatic HIV disease, using low-cost HAART prices would also significantly decrease hospital services utilization by HIV-infected patients, resulting in either health expenditure saving by cost deferral or freeing substantial resources for health care of non-HIV patients.
Patient Summary
Background
The number of cases of AIDS continues to increase worldwide; the disease is a major threat to humanity, with Africa facing the very worst problems. In South Africa alone there were 370,000 AIDS deaths in 2003. AIDS is caused by a type of retrovirus—the human immunodeficiency virus (HIV). Highly active antiretroviral treatment (HAART) is a treatment that uses a combination of three or more antiretroviral drugs that attack different parts of the virus. HAART is expensive, making it difficult for poor countries to provide treatment for all who need it. Prices are falling, however, and South Africa is one country where efforts are now being made to improve access to treatment.
Why Was This Study Done?
The cost-effectiveness of HAART has been studied in developed countries, but developing countries also need to know how much it is going to cost their health services if they introduce HAART, and whether there will be financial savings because of switching to a more effective treatment.
What Did the Researchers Do and Find?
During the study period (January 1995 to 31 December 2000), HAART was not available in the publicly funded South African health-care sector. The study, funded by the drug manufacturer Bristol-Myers Squibb, took place in HIV clinics affiliated with the University of Cape Town. The researchers compared the cost of services for 292 patients who were given HAART with the costs for a comparison group (with the same number of patients) who were not given any antiretroviral drugs. Twenty-seven patients in each group had AIDS; the others were HIV-infected but did not have AIDS. The researchers calculated costs per patient year (PPY) and per life-year gained (LYG), i.e., the total cost divided by the number of extra years the treated patients lived. Calculations were done separately for patients with AIDS and those without AIDS. Patients on HAART required fewer hospital admissions. Depending on how long the patient survived and the price of antiretrovirals, it cost less to treat the HAART patients with AIDS. For this group, the cost saving ranged from $219 to $2,116 (in U.S. dollars). For patients without AIDS, the cost of treatment (ranging from $597 to $1,772) was, by the South African standard of cost of living, affordable. However, it is expected that South Africa will soon be able to manufacture antiretroviral drugs locally and more cheaply. This would increase the amount saved by introducing HAART.
What Does This Mean?
HAART seems to be a more cost-effective way for South African hospitals to treat HIV infection than simply waiting for patients to come to hospital and then dealing with their symptoms. However, it should be noted that when a person is infected with HIV and becomes ill or dies from AIDS, it is not only hospitals that face costs. The patient, their family, and the country suffer financially. Effective treatment might also lower these “indirect” costs, but this was not an issue examined in this research.
Where Can I Find More Information Online?
For a comprehensive source of information on HIV/AIDS:
http://www.thebody.com
The site also includes a useful section on HAART:
http://www.thebody.com/Forums/AIDS/Infections/Archive/NewMedications/Q12178.html
For information about the global AIDS situation and the position in different countries:
http://www.unaids.org.
We are indebted to staff and patients of the HIV clinics from Somerset Hospital (NSH) and Groote Schuur Hospital (GSH). We are particularly indebted to Douglas Wilson, Rosalind Mayniar, Elizabeth Fielder, Mostafa Golayz, and Robin Hawkins at NSH and GSH. We are also grateful for the contributions of Jim Hanley at McGill University, Di McIntyre from the University of Cape Town, and Brian Gazzard from Chelsea and Westminster Hospital, United Kingdom. Study was funded by Secure the Future, Bristol-Myers Squibb, through an unrestricted educational grant. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Citation: Badri M, Maartens G, Mandalia S, Bekker L, Penrod JR, et al. (2006) Cost-effectiveness of highly active antiretroviral therapy in South Africa. PLoS Med 3(1): e4.
Abbreviations
ARTantiretroviral therapy
CIconfidence interval
HAARThighly active antiretroviral therapy
IQRinterquartile range
LYGlife-year gained
ORodds ratio
PPYper patient-year
WHOWorld Health Organization
==== Refs
References
UNAIDS HIV/AIDS epidemic in South Africa 2004 Available: http://www.unaids.org/en/geographical+area/by+country/south+africa.asp . Accessed 12 October 2005
Dorrington R Bourne D Bradshaw D Laubscher R Timaeus I The impact of HIV/AIDS on adult mortality in South Africa. Technical report for the South African Medical Research Council 2001 Available: http://www.mrc.ac.za/bod/complete.pdf . Accessed 12 October 2005
Floyd K Reid RA Wilkinson D Gilks C Admission trends in a rural South African Hospital during early years of the HIV epidemic JAMA 1999 282 1087 91 10493210
Hensher M Health sector impacts of HIV/AIDS. Key issues for planning 2001 Johannesburg (South Africa) Department of Health. Directorate Health Financing and Economics 45
Abt Associates Impending catastrophe revisited: An update on the HIV/AIDS epidemic in South Africa 2001 Parklands (South Africa) Henry J Kaiser Family Foundation 36
Freedberg KA Losina E Weinstein MC Paltiel AD Cohen CJ The cost effectiveness of combination antiretroviral therapy for HIV disease N Engl J Med 2001 344 824 831 11248160
Sendi PP Bucher HC Harr T Craig BA Schwietert M Cost-effectiveness of highly antiretroviral therapy in HIV-infected patients: SWISS HIV cohort study AIDS 1999 13 1115 1122 10397543
Moore R Cost-effectiveness of combination HIV therapy: 3 years later Pharmacoeconomics 2000 17 325 330 10947487
Miners AH Sabin CA Trueman P Youle M Mocroft A Assessing the cost-effectiveness of HAART for adults with HIV in England HIV Med 2001 2 52 58 11737376
Le Pen C Rozenbaum W Downs A Maurel F Lilliu H Effect of HAART on health status and hospital costs of severe HIV-infected patients: a modeling approach HIV Clin Trials 2001 2 136 145 11590522
Beck EJ Kennelly J McKevitt C Whitaker L Wadsworth J Changing use of hospital services and costs at a London AIDS referral centre, 1983–1989 AIDS 1994 8 367 377 8031516
Beck EJ Mandalia S Williams I Power A Newson R for the NPMS Steering Group (1999) Decreased morbidity and use of hospital services in English HIV infected individuals with increased uptake of anti-retroviral therapy 1996–1997 AIDS 13 2157 2164 10546870
Stoll M Class C Scuttle E Graf Von Der Sculenburg JM Schmidt RE Direct costs for the treatment of HIV-infection in a German cohort after introduction of HAART Eur J Med Res 2002 7 463 471 12568973
Torres RA Barr M Impact of combination therapy for HIV infection on inpatient census N Engl J Med 1997 336 1531 1532
Bozzette SA Joyce G McCaffrey DF Leibowitz AA Morton SC Expenditures for the care of HIV-infected patients in the era of highly active antiretroviral therapy N Engl J Med 2001 344 817 823 11248159
Beck EJ Tolley K Power A Mandalia S Rutter P for the NPMS Steering Group (1998) Use and cost of HIV service provision in England Pharmacoeconomics 14 639 652 10346416
Beck EJ Whitaker L Kennelly J McKevitt C Wadsworth J Changing presentation and survival, service utilization and costs for AIDS patients: insights from a London referral centre AIDS 1994 8 379 384 8031517
Cullinan K South Africa takes first steps to provide antiretrovirals Bull World Health Organ 2002 80 921 12481221
Medecins Sans Frontiers 2005 Medecins Sans Frontiers (2005) Untangling the web of price reductions A pricing guide for the purchase of ARVs for developing countries, 7th ed Available: http://www.accessmed-msf.org/documents/untanglingtheweb%207.pdf . Accessed 12 October 2005
UNAIDS Access to treatment and care. Fact sheet 2003. UNAIDS, 2003 2003 Available: http://www.redribbon.co.za/documents_v2/aaa_unai/fact/fsht9-_Treatment.doc . Accessed 21 October 2005
Department of Health ARV drug tender awarded 2005 Available: http://www.doh.gov.za/docs/dr/2005/pr/pr0303-f.html . Accessed 12 October 2005
Badri M Bekker LG Orrell C Pitt J Cilliers F Initiating highly active antiretroviral therapy in sub-Saharan Africa: An assessment of the revised World Health Organization scaling-up guidelines AIDS 2004 18 1159 1168 15166531
Badri M Wilson D Wood R Effect of highly active antiretroviral therapy on incidence of tuberculosis in South Africa: A cohort study Lancet 2002 359 2059 2064 12086758
World Health Organization Acquired immunodeficiency syndrome (AIDS). Interim proposal for a WHO staging system for HIV infection and disease Wkly Epidemiol Rec 1990 65 221 224 1974812
SAS Institute SAS OnlineDoc, version 8.02 [computer program] 2000 Cary (North Carolina) SAS Institute
Cape Metropolitan Council Levels of living in the Cape Metropolitan area, 2nd ed 1997 Cape Town (South Africa) Metropolitan Council 6
Beck EJ Mandalia S Gaudreault M Brewer C Zowall H The cost-effectiveness of highly active antiretroviral therapy, Canada 1991–2001 AIDS 2004 18 2411 2418 15622317
Beck EJ Miners AH Gazzard B Johnson M Miles A Effectiveness and efficiency in the delivery of HIV services: Economic and related considerations The effective management of HIV disease 2001 London Aesculapius Medical Press 113 138
Drummond MF O'Brien B Stoddart GL Torrance GW Methods for the economic evaluation of health care programmes, 2nd ed 1997 Oxford Oxford University Press 52 53
Beck EJ Miners A Tolley K The cost of HIV treatment and care. A global review Pharmacoeconomics 2001 19 13 39
Govender V McIntyre D Grimwood A Maartens G The costs and perceived quality of care for people living with HIV/AIDS in the Western Cape Province in South Africa. Small-Applied Research No. 14 2000 Bethesda (Maryland) Partnerships for Health Reform Project, Abt Associates Available: http://www.phrplus.org/Pubs/sar14.pdf . Accessed 12 October 2005
Statistics South Africa Historical PPI key indicators 2004 Available: http://www.statssa.gov.za/keyindicators/CPI/CPIHistory.pdf . Accessed 21 October 2005
US Federal Reserve Board Foreign exchange rates. South Africa historical rates (annual) 2005 Available: http://www.federalreserve.gov/releases/H10/hist/dat00_sf.htm . Accessed 21 October 2005
World Health Organization Scaling up antiretroviral therapy in resource-limited settings: Treatment guidelines for a public health approach. 2003 revision 2004 Geneva World Health Organization Available: http://www.who.int/hiv/pub/prev_care/en/arvrevision2003en.pdf . Accessed 12 October 2005
Parsonage M Neuberger H Discounting of health benefits Health Economics 1992 1 71 76 1342633
Ngalula J Urassa M Mwaluko G Isingo R Boerma T Health service use and household expenditure during terminal illness due to AIDS in rural Tanzania Trop Med Int Health 2002 7 877 877
Arthur G Bhatt SM Muhindi D Achiya GA Kariuki SM The changing impact of HIV/AIDS on Kenyatta National Hospital, Nairobi from 1988/89 through 1992 to 1997 AIDS 2000 14 1625 1631 10983650
Buve A AIDS and hospital bed occupancy: An overview Trop Med Intern Health 1997 2 136 139
Tembo G Friesan H Asiimwe-Okiror G Moser R Naamara W Bed occupancy due to HIV/AIDS in an urban hospital medical ward in Uganda AIDS 1994 8 1169 1171 7986417
Nunn P Gathua S Kibuga D Binge R Brindle R The impact of HIV on resource utilization by patients with tuberculosis in a tertiary referral hospital, Nairobi, Kenya Tuber Lung Dis 1993 74 273 279 8219180
Beck EJ Mandalia S Power A Easterbrook P Pozniak A Reduced HIV disease progression and mortality due to CART in English NPMS-HHC clinics [abstract]. XIII International AIDS Conference; 2000 9 July–14 July; Durban, South Africa 2000 for the NPMS-HHC Steering Group. Available: http://www.iac2000.org/abdetail.asp?ID=TuPeC3331 . Accessed 12 October 2005
Hore R Cross S Whiteside A The medical costs of AIDS in Zimbabwe Facing up to AIDS: The socio-economic impact in southern Africa 1993 New York St Martin's Press 241 261
Harling G Wood R Beck EJ A review of the efficiency of interventions in HIV infection, 1994–2004 Dis Manage Health Outcomes 2005 In press
Garber AM Culyer AJ Newhouse JP Advances in cost-effectiveness analysis of health interventions Handbook of health economics, volume 1A 2000 Amsterdam Elsevier 181 221
Bureau of Public Affairs, U.S. Department of State 2005 Department of State (2005) Background note South Africa Available: http://www.state.gov/r/pa/ei/bgn/2898.htm . Accessed 2005 April
Harries AD Nyangulu DS Hargreaves NJ Kaluwa O Salaniponi FM Preventing antiretroviral anarchy in Africa Lancet 2002 358 410 414
Orrell C Bangsberg DR Badri M Wood R Adherence is not a barrier to successful antiretroviral therapy in South Africa AIDS 2003 17 1 7 12478064
Laurent C Diakhate N Gueye NF Toure MA Sow PS The Senegalese government's highly antiretroviral therapy initiative: An 18-month follow-up study AIDS 2002 16 1363 1370 12131213
Weidle PJ Malamba S Mwebaze R Sozi C Rukundo G Assessment of a pilot antiretroviral therapy programme in Uganda: Patients' response, survival, and drug resistance Lancet 2002 360 34 40 12114039
Geffen N Natrass N Raubenheimer C The cost of HIV prevention and treatment interventions in South Africa. CSSR working paper no 28 2003 Cape Town (South Africa) University of Cape Town Centre for Social Science Research Available: http://www.hst.org.za/uploads/files/cost_sa.pdf . Accessed 12 October 2005
Boulle A Kenyon C Skordis J Wood R Exploring of the costs of a limited public sector antiretroviral treatment program in South Africa S Afr Med J 2002 92 811 817 12432807
Bell C Devarajan S Gersbach H 2003 Bell C, Devarajan S, Gersbach H (2003) The long-run economic costs of AIDS Theory and an application to South Africa Available: http://www1.worldbank.org/hiv_aids/docs/BeDeGe_BP_total2.pdf . Accessed 12 October 2005
|
16318413
|
PMC1298940
|
CC BY
|
2021-01-05 11:13:39
|
no
|
PLoS Med. 2006 Jan 6; 3(1):e4
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030004
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1631841010.1371/journal.pmed.0030006Research ArticleCancer BiologyGenetics/Genomics/Gene TherapyPharmacology/Drug DiscoveryDrugs and Adverse Drug ReactionsGeneticsOncologyEGF Receptor-Targeted Synthetic Double-Stranded RNA Eliminates Glioblastoma, Breast Cancer, and Adenocarcinoma Tumors in Mice Cancer-Specific Transfection of dsRNAShir Alexei
1
Ogris Manfred
2
Wagner Ernst
2
Levitzki Alexander
1
*1Department of Biological Chemistry, The Hebrew University of Jerusalem, Givat Ram, Jerusalem, Israel2Pharmaceutical Biology-Biotechnology, Department of Pharmacy, Ludwig-Maximilians-Universität, Munich, GermanyLiu Ed Academic EditorGenome Institute of SingaporeSingapore*To whom correspondence should be addressed. E-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
Author Contributions: AS, MO, EW, and AL designed the study. AS and MO performed the experiments. AS and AL analyzed the data. AS, MO, EW, and AL contributed to writing the paper.
1 2006 6 12 2005 3 1 e69 5 2005 29 9 2005 Copyright: © 2006 Shir et al.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Glioblastoma Multiforme-Treating a Deadly Tumor with Both Strands of RNA
New Approach to Combat Glioblastoma Shows Promise in Preclinical Studies
Background
Glioblastoma multiforme (GBM) is the most lethal form of brain cancer. With the available treatments, survival does not exceed 12–14 mo from the time of diagnosis. We describe a novel strategy to selectively induce the death of glioblastoma cells and other cancer cells that over-express the EGF receptor. Using a non-viral delivery vector that homes to the EGF receptor, we target synthetic anti-proliferative dsRNA (polyinosine-cytosine [poly IC]), a strong activator of apoptosis, selectively to cancer cells.
Methods and Findings
Poly IC was delivered by means of a non-viral vector: 25kDa polyethylenimine-polyethyleneglycol-EGF (PEI25-PEG-EGF). EGFR-targeted poly IC induced rapid apoptosis in the target cells in vitro and in vivo. Expression of several cytokines and “bystander killing” of untransfected tumor cells was detected in vitro and in vivo. Intra-tumoral delivery of the EGFR-targeted poly IC induced the complete regression of pre-established intracranial tumors in nude mice, with no obvious adverse toxic effects on normal brain tissue. A year after treatment completion the treated mice remain cancer-free and healthy. Similarly, non-viral delivery of poly IC completely eliminated pre-established breast cancer and adenocarcinoma xenografts derived from EGFR over-expressing cancer cell lines, suggesting that the strategy is applicable to other EGFR-over-expressing tumors.
Conclusion
The strategy described has yielded an effective treatment of EGFR over-expressing GBM in an animal model. If this strategy is translated successfully to the clinical setting, it may actually offer help to GBM patients. Moreover the elimination of two additional EGFR over-expressing cancers in vivo suggests that in principle this strategy can be applied to treat other tumors that over-express EGFR.
A non-viral vector delivers double-stranded RNA predominantly to tumor cells overexpressing EGFR. This leads to apoptosis in tumor cells and eliminates established tumors in mice.
==== Body
Introduction
Glioblastoma multiforme (GBM), a brain cancer, is one of the deadliest human diseases, and cannot be cured by any therapy available today. The localization of GBM in the central nervous system and the very solid structure of this tumor renders it almost impermeable to large particles, such as viral vectors [1]. A major challenge in the treatment of GBM is to kill the accessible cancer cells on the surface of the tumor more rapidly than the rate of replication of the cells. Otherwise, the unexposed, internal cells can replicate and compensate for the cells that have just been eliminated. Thus, an effective treatment for GBM must incorporate the following features: (a) high selectivity and safety, to avoid damage to non-cancerous brain tissue; (b) rapid and efficient cell killing, preferably by simultaneous activation of multiple killing mechanisms. The simultaneous activation of multiple killing pathways will ensure tumor cell death, even if one or several pathways are inactive; and, (c) inhibition of the growth or killing of neighboring, unexposed tumor cells. This “bystander effect” should assist in eliminating the tumor before it can re-grow. It should also inhibit the growth of any tumor cells that may have a different phenotype from the targeted cells and are not themselves targeted by the treatment, including cancer stem cells.
In an attempt to meet all these demands in one treatment, we have taken advantage of the frequent (50%–70%) over-expression of epidermal growth factor receptor (EGFR) in GBM [2]. We have attached synthetic, double-stranded RNA (dsRNA) to a non-viral vector that can home in on EGFR. The dsRNA is selectively introduced into the cancer cells via receptor-mediated endocytosis. Double-stranded RNA, frequently expressed in cells infected with viruses, activates a number of pro-apoptotic processes simultaneously. These include the dsRNA dependent protein kinase (PKR) and 2,5- oligo-A synthetase, both of which turn off protein synthesis [3]. Double-stranded RNA also activates p38 and JNK, and stimulates the synthesis of pro-apoptotic proteins, such as IRF3-DRAF1 and NFκB [3–5]. These dsRNA-induced mechanisms efficiently kill infected cells and induce expression of anti-proliferative cytokines from the interferon family, thereby preventing spread of the virus [4].
In order to specifically introduce poly IC into EGFR over-expressing cells, we utilized polyethylenimine (25 kDa)-polyethylene-glycol-mEGF (PEI25-PEG-EGF) complexes [6,7]. We expected this approach to be highly selective, because the number of EGFRs on tumor cells is 10–100 times higher than that on non-tumor cells [2]. PEI25-PEG-EGF conjugates are significantly safer than replication-deficient or replication-competent viruses, in terms of immunotoxic reactions, inadvertent recombination and viral replication in healthy cells. Cell death was expected to be fast, because dsRNA activates cell killing mechanisms within minutes of entering the cell. Finally, induction of interferons, clinically used against GBM, was expected to exert a bystander effect and inhibit the growth of adjacent, untransfected tumor cells.
Methods
Reagents and Assays
Poly IC was obtained from Sigma (Rehovot, Israel). It was dissolved in DEPC-treated double-distilled H2O. The polyethylenimine (PEI), PEI25, branched and succinimidyl 3-(2-pyridyldithio) propionate (SPDP) were purchased from Sigma-Aldrich (Munich, Germany). NHS-PEG-MAL (MW = 3400) was obtained from Nektar Therapeutics (Huntsville, Alabama, United States) and the recombinant mouse EGF (mEGF) from Pepro Tech EC Ltd. (London, United Kingdom). The PEI content of the conjugate was determined spectrophotometrically by TNBS assay at 405 nm. The amount of dithiopyridine linkers in PEI was determined after reduction of an aliquot with dithiothreitol (DTT) followed by absorption measurement of released pyridine-2-thione at 343 nm.
The molar ratio of mEGF: dithiopyridine was determined spectrophotometrically at 280 and 340 nm. The amount of dithiopyridine was determined as described [6,7]. The yield of mEGF (mg) was calculated in two equations. Equation 1: A280 (a) = A340 with DTT × 5.1/8.1. Equation 2: A280 revised = A280−A280 (a). The result of equation 2 was the amount of mEGF in mg.
The Ellman assay was used for the determination of the mercapto groups in mEGF-SH. Liquid chromatography of conjugates was performed with the ÄKTA basic system from Amersham Biosciences (Little Chalfont, United Kingdom).
Melittin (Mel) (D-Mel-SH; 280 = 5570, MW = 2893.6) was purchased from IRIS Biotech GmbH (Marktredwitz, Germany).
All other chemicals were purchased from Sigma-Aldrich.
Fluorescence Microscopy
Poly IC was labeled with the Fluorescein ULS labeling kit (Fermentas, Hanover, Maryland, United States) at 1 unit of fluorescein per 1 μg of poly IC according to the manufacturer's instructions and then condensed with the appropriate PEI conjugate. Cells were incubated with the complexes in DMEM/FCS for 4 h at 37 °C and washed twice with PBS. Cells were viewed on a Zeiss confocal microscope. Green fluorescence was viewed with filter sets appropriate for fluorescein.
In Vitro Apoptosis Detection
Cells were seeded into 24-well plates at a density of 10,000 in 1 ml of medium per well and grown overnight. After appropriate treatment cells were washed, fixed and stained using the Annexin-V-Biotin kit (Roche, Basel, Switzerland) according to the manufacturer's instructions. In addition apoptotic death was also determined by TUNEL assay using the In situ Cell Death Detection kit (Roche). The brown colored apoptotic cells were visualized in a microscope, counted (6 fields per sample), and photographed using digital camera.
In Vitro Bystander Effect
For this assay, 500,000 U87MGwtEGFR cells were seeded onto 6-cm plates, grown overnight in 2 ml of medium, and transfected with 1 μg/ml poly IC using PEI-PEG-EGF+PEI-Mel complexes. Medium was collected at 24 h after transfection. U87MG and U87MGΔEGFR “indicator” cells were seeded in duplicate in 96-well plates (4,000 cells/well) and grown overnight in 200 μl of medium. Then, 100 μl of medium was then replaced by the medium collected from the transfected (+poly IC) or untransfected (−poly IC) U87MGwtEGFR cells. Where indicated the medium was pre-incubated for 1 h at room temperature, with neutralizing polyclonal anti-IFN-α antibody (1:500, Santa Cruz Biotechnology, Santa Cruz, California, United States). In NT samples, medium was not replaced. Growth inhibition was examined 48 h after medium exchange.
Effect of Targeted Poly IC on 10-d-Old Intracranial Tumor Models
For this assay, 10,000 U87MGwtEGFR cells per animal were stereotactically implanted into the brains of 35 nude mice as described [8]. Five animals were sacrificed 10 d later to evaluate the sizes of the tumors [8]. Then, 200 μl of Alzet (Alzet, Cupertino, California, United States) osmotic micropumps with intra-tumoral catheters were installed in 20 mice. Ten mice received (poly IC)PEI-PEG-EGF+PEI-Mel complex dissolved in HBG buffer [9] at 0.1 μg poly IC/μl buffer (the dose of poly IC was 0.8 μg/h or 19.2 μg/d), and ten others received equivalent doses of PEI-PEG-EGF+PEI-Mel complexes only (no poly IC). Ten remaining control animals did not receive any treatment. The pumps were replaced twice every 24 h. At day 20 after cell implantation (10 d after treatment initiation), five animals from each group were sacrificed to evaluate tumor size as described [8]. Five other animals in each group were retained for survival analysis [8]. All animal experiments were conducted in accordance with the Hebrew University guidelines for the care of laboratory animals.
In Vivo Bystander Effect
For this assay, 5,000 U87MGwtEGFR cells were mixed with 5,000 U87MGΔEGFR cells in 5 μl of PBS per animal. The cells were stereotactically implanted into the brains of 14 nude mice as described [8]. 10 d later, 200-μl Alzet osmotic micropumps with intra-tumoral catheters were installed in 12 mice. Six mice received (poly IC)PEI-PEG-EGF+PEI-Mel (poly IC/conjugate) complexes dissolved in HBG buffer at the indicated dose (two animals for each dose), and six others received an equivalent dose of PEI-PEG-EGF+PEI-Mel complexes without poly IC (Conjugate only). The pumps were replaced twice every 24 h. Survival of the animals was analyzed as above.
In Vivo Apoptosis Detection
For this assay, 10,000 U87MGwtEGFR cells per animal were stereotactically implanted into the brains of 12 mice as above. 14 d later, 200-μl Alzet osmotic micropumps with intra-tumoral catheters were installed in six U87MGwtEGFR-bearing mice. Three animals received formulated poly IC (19.2 μg total per animal for 24 h), and three mice received complexes without poly IC. Six remaining animals received no treatment. Immediately after termination of infusion (15 d after cell implantation), animals were sacrificed, the brains were fixed with 4% formalin, embedded in paraffin, and ultra-thin slices were prepared. Slices were then de-paraffinized and analyzed by fluorescent immunohistochemistry to determine apoptosis with the Cell Death Detection kit-TMR Red, Roche (red fluorescence) and EGFR expression with FITC conjugated EGFR (internal domain) antibody (green fluorescence), Biosource (Camarillo, California, United States).
Synthesis of Mel-PEI25-PEG-mEGF
The synthesis of mEGF-PEG-PEI25 is described elsewhere [6,7]. mEGF-PEG-PEI25 (83 nmol PEI) was mixed with SPDP (664 nmol in 100% ethanol) under argon. After 3 h at room temperature the mixture of about 2 ml was loaded on a gel filtration column (Sephadex G25 superfine; HR10/30; 20mM HEPES [pH 7.1], 0.5M NaCl; Amersham Biosciences). The purified PDP-functionalized conjugate (5 ml) containing 309 nmol of PDP was concentrated to 1.5 ml by speed vac. For the reaction with Melittin 464 nmol of Mel was weighed out and dissolved in 0.5 ml of 0.5 M NaCl, 100mM HEPES [pH 7.4] degassed with argon. Both components were mixed under argon. After 20 h at room temperature mEGF-PEG-PEI25-Mel was purified by gel filtration. To gel filtrate the conjugate, a Superdex 75 prep grade column 10/30, conditioned with PEI25br (10 mg PEI25/60 ml gel material) was used. After dialysis overnight (MWCO 14000; Visking type 27/32; Roth, Karlsruhe, Germany) against HBS 6 ml of mEGF-PEG-PEI25-Mel conjugate were obtained; these contained 66 nmol PEI (1.64 mg), 350 nmol Mel, and 70 nmol EGF.
A431 Xenograft Treatment
Two million A431 cells were implanted subcutaneously into the left flanks of 15 nude CD-1 mice. When palpable tumors (average size 10.1 mm3) had developed, animals were divided into three groups of five animals. The complexes were delivered by intra-tumoral injection twice per day for 6 d. Five mice received poly IC-Melittin-PEI25-PEG-EGF (poly IC-MPPE) complex dissolved in HBG buffer at 0.1 μg poly IC/μl buffer (the dose of poly IC was 15 μg/injection or 30 μg/d), and five others received equivalent doses of MPPE complexes only (no poly IC). Remaining animals did not receive any treatment. Length (a) and width (b) of the tumors were measured daily with caliper and the tumor size was calculated as ab2/2. Control animals were sacrificed at day 33 after treatment initiation. Poly IC treated animals were left for follow up study to detect recurrence of the tumors.
MDA-MB-468 Xenograft Treatment
Two million MDA-MB-468 cells were implanted over the mammary fat pad of 15 SCID/non-obese diabetic (NOD) mice. When palpable tumors (average size 9.8 mm3) had developed, animals were divided into three groups of five animals each. The complexes were delivered by intra-tumoral injection once per day for 6 d. Five mice received (poly IC)MPPE complex dissolved in HBG buffer at 0.1 μg poly IC/μl buffer (the dose of poly IC was 20 μg/d), and five others received an equivalent doses of MPPE complexes only (no poly IC). Tumor size was calculated as above. Control animals were sacrificed at day 33 after treatment initiation. Poly IC-treated animals were left for follow-up study to detect recurrence of the tumors.
Results
Fast and Selective Killing of EGFR Over-Expressing GBM Cells In Vitro
The PEI25-PEG-EGF complexes efficiently delivered poly IC, killing up to 85% of U87MGwtEGFR cells, which over-express EGFR (∼1 × 106 receptors [10]), within 1 h of transfection (Figure 1A). At poly IC concentrations of up to 10 μg/ml, no significant effect was observed on the parental U87MG cells, which express 100,000 of EGFR per cell [10], on cells that over-express the mutated Δ(2–7)EGFR (U87MGΔEGFR), or on glioma cells lacking the EGFR (U138MG [11]) (Figure 1A). At the high concentration of 20 μg/ml poly IC, the survival of U87MG and U87MGΔEGFR cells, was inhibited by 20%. U138MG cells, which completely lack EGFR, were not inhibited at all. The killing effect on U87MGwtEGFR cells was enhanced 8- to 10-fold when PEI25-PEG-EGF was partially replaced with the polyethylenimine (2 kDa)-Melittin conjugate PEI2-Melittin (Figure 1A), and more than 95% of U87MGwtEGFR died within an hour of transfection, again, with no effect on the other glioma cell lines (Figure 1A). Melittin is a bee venom peptide that facilitates the release of nucleic acids from the endosome into the cytoplasm [9,12], thus enhancing the release of poly IC from the endocytosed vesicle. No significant toxic effect of the complexes without poly IC on any of the cell lines was detected (Figure S1). Annexin V and TUNEL staining showed that the majority (70%–90%) of the U87MGwtEGFR cells died by apoptosis within 1 h of transfection (Figure 1B and 1C). Although fast apoptotic death is not common, it has occasionally been detected in other cell lines where, like here, a number of pro-apoptotic pathways are activated simultaneously [13,14].
Figure 1 (poly IC)PEI-PEG-EGF(+PEI-Mel) Complexes Selectively Kill U87MGwtEGFR Cells
(A) Cells were seeded in duplicate onto a 96-well plate at a density of 5,000 cells in 0.2 ml of medium per well and grown overnight. Cells were then transfected as described [6,9] with poly IC at the indicated concentrations using either PEI-PEG-EGF or PEI-PEG-EGF+PEI-Mel (w/w ratio PEI-PEG-EGF:PEI-Mel = 1:10) complexes. Viability was measured by the CellTiter-Glo Luminescent Cell Viability Assay (Promega) according to the manufacturer's instructions, at 1 h after transfection.
(B and C) Fast induction of apoptosis by (poly IC)PEI-PEG-EGF+PEI-Mel complexes. Apoptotic death was detected 1 h after transfection by Annexin (B) and TUNEL (C) assays as described in Methods.
Distribution of the Formulated Poly IC in the Cells
In order to verify selective entrance of the complexes into the target U87MGwtEGFR cells and its release from endosomes fluorescent labeling of poly IC was performed (Methods). U87MGwtEGFR and U87MG cells were transfected with the labeled poly IC/PEI-PEG-EGF complex either in presence or absence of PEI-Mel (Figure 2). Figure 2A shows efficient transfection of the target U87MGwtEGFR cells and virtually no signal in U87MG cells. To examine the intracellular distribution, U87MGwtEGFR cells incubated with fluorescently labeled poly IC/PEI-PEG25-EGF or poly IC/PEI-PEG25-EGF+PEI-Mel complexes, were washed, and viewed live to rule out fixation artifacts, using confocal microscopy (Figure 2B). After 4 h of incubation, poly IC/PEI-PEG25-EGF complexes appeared in a punctate intracellular pattern, suggesting entrapment within vesicles (Figure 2B). In contrast, poly IC/PEI-PEG25-EGF+PEI-Mel complexes showed fluorescence dispersed throughout the cytoplasm (Figure 2B). The cytoplasmic fluorescence of the Melittin-containing complexes suggests that Melittin indeed facilitated release of the complex into the cytoplasm, by lysis of intracellular vesicles.
Figure 2 Distribution of the Complexes
Cells were seeded into 6-cm plates at a density of 300,000 cells in 2 ml of medium per plate and grown overnight. Cells were then transfected with the fluorescently labeled poly IC (5 μg/ml) using the indicated conjugates. After 4 h, cells were analyzed by fluorescent microscopy (Methods) for selectivity of the transfection (A) and intracellular distribution of the complex in the U87MGwtEGFR cells (B).
Involvement of PKR in Cell Death
The fast killing effect of the complexes was significantly inhibited by 2-aminopurine (2-AP), a potent inhibitor of PKR [15], suggesting that this effect is largely mediated by PKR (Figure 3A). At later time points, 48 and 72 h after transfection, the protective effect of 2-AP was significantly less pronounced, suggesting that additional cell killing mechanisms were activated.
Figure 3 Cell Killing Mechanisms
(A) Protection of the cells by 2-AP. Cells were grown as in Figure 1A. Cells were then transfected with poly IC at the indicated concentrations using PEI-PEG-EGF+PEI-Mel complexes. Where indicated, 2-AP (5 mM) was added 18 h before transfection and the medium was replaced every 24 h with medium containing fresh 2-AP. Viability was measured by the Methylene Blue assay [8].
(B) In vitro bystander effect. U87MGwtEGFR cells were grown and transfected as described in Methods. U87MG and U87MGΔEGFR “indicator” cells were grown in duplicates in 96-well plates. Medium of the “indicator cells” was then partially replaced by the medium collected from the transfected (+poly IC) or untransfected (-poly IC) U87MGwtEGFR cells (Methods). Where indicated the medium was preincubated with neutralizing polyclonal anti IFNα antibody. In NT samples, medium was not replaced.
(C) A total of 4,000 U87MGwtEGFR and U87MGΔEGFR cells were seeded in duplicate onto a 96-well plate at the indicated ratios and grown overnight. Cells were then transfected with poly IC at the indicated concentrations using PEI-PEG-EGF+PEI-Mel conjugates. Cell survival was measured by the Methylene Blue assay 96 h after transfection.
In Vitro Bystander Effect
We next investigated whether the poly IC-transfected cells exerted a bystander effect on untransfected cells (Methods). Growth medium from poly IC-transfected U87MGwtEGFR cells inhibited the growth of U87MG and U87MGΔEGFR cells (Figure 3B), which are themselves insensitive to the (poly IC)PEI-PEG-EGF+PEI-Mel complex (Figure 1A). The bystander effect was less pronounced when U87MGwtEGFR cells were transfected with targeted poly IC at higher concentrations (Figure S2), probably because the transfected cells died before they could exert the full effect. The bystander effect was further confirmed by co-culture of U87MGwtEGFR and U87MGΔEGFR cells in the same wells (Figure 3C). A large fraction of the U87MGΔEGFR cells were killed, although they are themselves almost insensitive to the poly IC complex.
Double-stranded RNA is known to induce the expression of interferons and other anti-proliferative cytokines. To investigate whether the bystander effect was caused by interferon, we incubated medium from transfected U87MGwtEGFR cells with neutralizing IFNα antibody. This pre-incubation reduced the bystander effect on all cell lines by 20%–30% (Figure 3B), indicating that IFN-α was indeed responsible for part of the bystander effect. The partial reduction in the bystander effect suggests that additional cytokines, other than IFN-α, were also expressed. Using ELISA (IBL, Hamburg, Germany), we confirmed the presence of IFNα in growth medium from poly IC-transfected cells. Up to 4 pg/ml IFNα were generated in the medium of 500,000 U87MGwtEGFR cells transfected with 2.5 μg/ml of poly IC, but not in the medium of U87MG or U87MGΔEGFR cells treated with the same complex. No IFN-α was detected in medium from cells transfected with poly IC at higher concentrations, suggesting that these cells were killed before they could secrete the cytokine. IFN-α was also generated in vivo (> 600 pg/g of tumor tissue), specifically in U87MGwtEGFR tumor xenografts (Figure 4). Using a Human Cytokine Microarray (RayBiotech, Norcross, Georgia, United States), we sought to identify additional cytokines in the medium of poly IC-transfected cells. Both GROα [16] and IP-10 [17] were detected. The production of IP-10 and Gro-alpha at high concentrations in vitro and selectively in tumors in vivo, was confirmed by cytokine specific ELISAs (Figure 4). These cytokines are chemokines responsible for the recruitment of T-cells to area of expression [16,17].
Figure 4 Targeted Poly IC Induces Expression of Cytokines
For this assay, 10,000 U87MGwtEGFR cells were implanted into the brain of nude mice. After 15 d, poly IC was injected with Alzet micropumps for 24 h at 0.8 μg/hr. After 48 h, animals were sacrificed and the brains were extracted. Tumors (average weight 17.6 mg) and surrounding brain tissue (average weight 253 mg) were resuspended in Tris-HCl buffer (100 mM Tris, [pH 8.1] with 1% Triton X-100) at 1:10 w/v. Samples were homogenized on ice by sonication, triturated through 19 gauge needles and spun at 20,000 × g. Protease inhibitors were added to supernatant, which was then subjected to cytokine-specific ELISA.
Formulated Poly IC Eliminates Intracranial GBM Models in Mice
These encouraging results led us to test the EGFR-targeted poly IC strategy in vivo. Ten thousand U87MGwtEGFR cells were implanted into the brains of nude mice as described (Methods) and the tumors left to grow for 10 d. During this period, visible tumors developed (Figure 5A). On day 10 after cell implantation, (poly IC)PEI-PEG-EGF+PEI-Mel complexes were delivered directly into the tumors at a constant rate for 3 d, using Alzet osmotic micropumps. On day 20 after cell implantation (7 d after the end of treatment), the tumors had disappeared completely from (poly IC)PEI-PEG-EGF+PEI-Mel-treated animals, while the tumors continued to grow in untreated animals until reaching 36.44 mm3 (Figure 5A). Animals that were treated with PEI-PEG-EGF+PEI-Mel alone (no poly IC), survived for no longer than 32 d, as did untreated animals (Figure 5B). In contrast, the (poly IC)PEI-PEG-EGF+PEI-Mel-treated animals are still alive on the day of submission of this manuscript (more than a year), and show no signs of increased intracranial pressure. When Melittin was omitted, the mice died from GBM 57–59 d after cell implantation (unpublished data). Histopathologic examination of the brains of the poly IC-treated mice at day 7 after the end of treatment did not reveal any residual tumors (Figures 5A and 6A), whereas the control animals had large tumors (Figures 5A and 6A). Pathological analysis of the brains at 24 h after treatment initiation showed increased gliosis in the brains of the animals of all experimental groups (Figure 6A). This was probably caused by the growing tumors in the brains. We also detected low infiltration of macrophages into the tumors in all groups (Figure 6A). None of the animals treated with poly IC showed any additional histopathological signs of toxicity or brain tissue damage either at 24 h after treatment initiation or at 7 d after the end of the treatment (Figures 5A and 6A). Similarly, we did not detect any significant increase in infiltration of immune cells into poly IC treated tumors as compared with the untreated tumors (Figure 6A).
Figure 5 Targeted Poly IC Eliminates Intracranial GBM Models
(A and B) (poly IC)PEI-PEG-EGF+PEI-Mel complexes eliminate U87MGwtEGFR xenografts. Intracranial U87MGwtEGFR tumors were established and treated with formulated poly IC as described in Methods. (A) Sizes of the tumors before and after the treatment. (B) Survival of the animals (Methods).
(C) In vivo bystander effect. A mixture of 5,000 U87MGwtEGFR and 5,000 U87MGΔEGFR cells was implanted into the brain of nude mice as described in Methods. The mixed tumors were treated with formulated poly IC (poly IC-/conjugate) at the indicated doses, and control animals received equivalent doses of PEI-PEG-EGF+PEI-Mel complexes without poly IC (Conjugate only). Survival of the animals was analyzed as above. Mice receiving 0.2μg/h of poly IC die up to day 75, whereas those receiving 0.4 μg poly IC/h and 0.8 μg poly IC/h are still alive.
Figure 6 In Vivo Selectivity of the Approach
(A) Pathological analysis of the brains. Intracranial tumors were established and treated as described in Methods. After 24 h, animals were sacrificed and ultrathin slices of the brains were prepared. Slices were stained with H & E and analyzed by light microscopy (20× magnification) for development of pathological signs and immune cell infiltration (yellow arrow). Bottom panel shows the example of slices from the brains treated with (poly IC)PEI-PEG-EGF+PEI-Mel in the experiment described in Figure 5A (7 d post-treatment).
(B) Poly IC induces apoptosis in intracranial xenografts. Intracranial tumors were established and treated as described in Methods. Apoptotic death was detected using Cell Death Detection kit-TMR Red (Methods). White dashed lines represent borders of the tumors.
Bystander Effect In Vivo
In the clinical setting, some GBM cells over-express wt EGFR, whereas neighboring cells express lower levels of EGFR or over-express truncated EGFR (ΔEGFR) [2]. As described above, in culture the EGFR-targeted poly IC-transfected cells induced the killing of neighboring, untransfected cells (Figure 3B and 3C). We therefore examined bystander killing in vivo (Methods). We implanted a mixture of 5000 U87MGwtEGFR and 5000 U87MGΔEGFR cells in mice brains (Figure 5C). The mixed tumors that developed 10 d later were treated with the targeted poly IC at various doses (Figure 5C). Control animals survived for no more than 34 d, while mice that received the lowest dose of poly IC complex (0.2 μg/h) survived more than 2 times longer. Animals that received higher doses of formulated poly IC (0.4 μg/h and 0.8 μg/h) are alive and well on the day of submission of this paper (day 297+). This very encouraging result is probably due to the strong bystander effect induced by the slow delivery of the formulated poly IC. Elimination of the mixed tumor was not caused by immune reaction since we did not detect increased infiltration of immune cells into the poly IC treated tumors (Figure 6A).
Formulated Poly IC Induces Apoptosis in U87MGwtEGFR Cells In Vivo
We next examined the mode of U87MGwtEGFR cell death induced by the targeted poly IC in vivo. U87MGwtEGFR cells were injected into mice and the established tumors were treated with formulated poly IC. EGFR expression was monitored with FITC-conjugated EGFR antibody (green fluorescence). Apoptosis was monitored with the TMR Red Cell Death Kit (Roche). Apoptosis occurred exclusively in cells over-expressing the EGFR (Figure 6B). No significant red signal was observed in the surrounding tissue, demonstrating the high selectivity and low toxicity of the treatment regimen.
Effect of Formulated Poly IC on 15-d-Old Intracranial Xenografts
Although 10-d tumors used in the survival experiments are clinically relevant to newly diagnosed GBM in humans (R. Catane and R. Pfeffer, Division of Neuro-oncology, Tel Hashomer/Sheba Hospital, personal communication), we examined whether formulated poly IC is effective against much larger tumors (Figure 5A). For this purpose we synthesized Mel-PEI25-PEG-EGF (MPPE, Methods). Although not observed, there is a possibility of PEI-Mel dissociation from (poly IC)PEI-PEG-EGF+PEI-Mel complexes. This may result in decreased release of poly IC to cytoplasm. When both Melittin and EGF are covalently bound to the same PEI molecule, the vector should be more efficient and even less toxic, because Melittin is now covalently linked to the complex, and therefore there is no possibility of PEI-Mel leakage. It was also observed that PEI25 conjugates are significantly smaller in size than PEI2 or PEI2+PEI25 conjugates (M.O., unpublished data), leading to higher in vivo transfection efficiency.
In order to obtain larger tumors, the intracranially implanted 10,000 U87MGwtEGFR cells were left to grow for 15 d. The resulted tumors were 13–15 times bigger than the 10-d-old tumors (Figures 5A and 7A). (poly IC)Mel-PEI-PEG-EGF and (poly IC)MPPE complexes were delivered directly into the tumors at a constant rate for 3, 4, and 5 d, using Alzet osmotic micropumps. Animals that were treated with MPPE alone (no poly IC), survived for no longer than 33 d, as did untreated animals (Figure 7A). In contrast, the (poly IC)MPPE-treated animals in all groups are still alive on the day of submission of this manuscript (day 244 +), and show no symptoms of GBM.
Figure 7 Targeted Poly IC Eliminates Three Types of EGFR Over-Expressing Tumors
(A) (poly IC)Mel-PEI-PEG-EGF (MPPE) complexes prolong survival of mice bearing large intracranial U87MGwtEGFR xenografts. Cells were implanted into the brains of 16 mice. 15 d later, two animals were sacrificed to measure the tumors (Methods). Other animals received the indicated treatments. The daily doses of complexes were similar to the doses in Figure 3A and 3B. Survival of the animals was analyzed as above.
(B) Formulated poly IC selectively kills A431 and MDA-MB-468 cells. Cells were seeded in duplicate onto a 96-well plate at a density of 5,000 cells in 0.2 ml of medium per well and grown overnight. Cells were then transfected as described [6,9] with poly IC at the indicated concentrations using either MPPE or PEI-PEG-EGF+PEI-Mel. Cell survival was measured by the Methylene blue assay at 48 h after transfection.
(C) (poly IC)MPPE complexes eliminate A431 and MDA-MB-468 xenografts in mice. A431 and MDA-MB-468 tumors were established and treated with formulated poly IC as described in Methods. Tumors were measured daily. Control animals were euthanized at day 33 after treatment initiation. Poly IC treated mice were kept alive to detect possible late recurrence of the tumors.
Formulated Poly IC Destroys Other EGFR Over-Expressing Cancers In Vitro and In Vivo
We next examined whether our strategy could be implemented for the treatment of other EGFR over-expressing cancers. We used the A431 (vulval carcinoma, expressing ∼2 × 106 EGF receptors per cell [18]) and MDA-MB-468 (breast cancer, similar EGFR expression and EGFR responsiveness to A-431 cells [19]), both non-engineered cell lines. Figure 7B shows targeted poly IC induced killing of A431 and MDA-MB-468 cells, whereas the control U87MG cells remain unharmed. Figure 7B also demonstrates the enhanced efficacy of the all-in-one vector MPPE as compared to PEI-PEG-EGF+PEI-Mel combination formula.
Encouraged by these results, we examined the efficiency of our strategy on models of these cancers in vivo. A431 cells were implanted subcutaneously into nude mice as described (Methods) and the tumors left to grow for 10 d. During this period, tumors of average size 10.1 mm3 were developed (Figure 7C). After tumor establishment, (poly IC)MPPE complexes were injected directly into the tumors twice per day for 6 d at 15 μg of poly IC/mouse/injection (30 μg/d) (Methods). On day 10 after treatment initiation, the tumors had disappeared completely from (poly IC)MPPE treated animals, while the tumors continued to grow in untreated animals, reaching up to139 mm3 (Figure 7C).
MDA-MB-468 cells were implanted over mammary fat pads of female SCID mice and the tumors left to grow for 14 d. During this period, tumors of average size 9.7 mm3 developed (Figure 7C). (poly IC)MPPE complexes were injected directly into the tumors for 6 d. On day 8 after treatment initiation, the tumors had disappeared completely from (poly IC)MPPE animals, while the tumors continued to grow in untreated animals, reaching 63 mm3 (Figure 7C).
We did not detect recurrence of any of the tumors in the follow up study (day 42+ after treatment initiation).
Discussion
Our results suggest that the EGFR-targeted delivery of poly IC can be implemented in the clinical treatment of GBM, for which current therapies are essentially ineffective. The therapeutic approach described here incorporates rapid and efficient killing of tumor cells by multiple mechanisms, a strong bystander effect, along with high selectivity. The localization of GBM in the central nervous system makes it an attractive candidate for local therapy by slow, constant, intra-tumoral delivery of the complexes described here (Figure 5). EGFR-targeted delivery of poly IC should be especially effective for the small tumors remaining after surgery, but could in principle be the first line of treatment, considering the results reported here. We have previously described an approach for GBM treatment using viral vectors to induce the in situ production of cancer-specific dsRNA within the cancer cell [8]. The EGFR-targeted synthetic dsRNA approach we describe now seems to be far superior, probably because we have achieved rapid delivery of a large dose of longer dsRNA into the cancer cell. This leads to a very fast response and induction of a bystander effect so that the tumor is demolished more rapidly than it can re-grow. Although many of the immune cells are still present in the nude mice we used, no significant immune reaction against the poly IC treated tumor was observed (Figure 6A). Most likely absence of the immune response is due to the fast elimination of the tumor. An immune reaction might be induced in patients following treatment that will take significantly longer periods of time. Nevertheless, expression of immunoactive IFN-α and T-cell chemokines, IP-10 and Gro-α selectively in the tumor (Figure 4) should drive the immune response specifically against GBM, reducing potential toxic effects on surrounding brain tissue.
To the best of our knowledge, the strategy described in this study has yielded the most effective treatment of EGFR over-expressing GBM reported so far, in an animal model. The therapeutic strategy described here differs significantly from other EGFR targeted agents aimed at GBM, such as erlotinib, gefitinib, [20] and anti-EGFR antibodies [21]. These agents inhibit the activity of the receptor and its downstream signaling. The response of GBM to gefitinib and erlotinib in GBM is weak [22]. Furthermore, the response of non-small-cell-lung cancer to either drug as a single agent is limited to tumors in which the EGFR harbor mutations in the kinase domain, where the EGFR is a survival element, and is usually transient [23]. Mutations (extracellular) in the EGFR receptor or frequent heterogeneity of GBM where only part of the cells over-express EGFR will result in only partial response to gefitinib and erlotinib as indeed observed in the recent clinical trials [22]. In the case of anti-EGFR therapy, the response seems not to be related to whether the EGFR is mutated but these drugs are also not life saving in patients with tumors involving EGFR [24]. The strategy described here is different from those currently implemented therapies because the EGFR is utilized to target a vehicle that delivers a strong pro-apoptotic molecule, namely dsRNA. Due to the strong bystander effect induced by the massive amount of dsRNA on neighboring cancer cells that were not targeted themselves, the therapy described here can eradicate EGFR over-expressing tumors even when only half of the cells over-express wt EGFR (Figure 5C). The bystander effect is mediated at least partially by IFN-α (Figure 3B). This cytokine is clinically used against various cancers including GBM. Since IFN-α kills mainly fast proliferating cells, normal cells should not be affected. Indeed, we did not detect any pathology in normal brain tissue during or after the therapy (Figure 6A).
Another important advantages of the therapy described here are the fast cell killing (Figure 1) and activation of several growth inhibitory pathways simultaneously (Figures 3A and 4). This form of therapy is likely to prevent and/or overcome a potential mutation in one of the anti-proliferative proteins and development of resistance to the therapy. This is especially relevant for cancers with deficient PKR activity [25], because poly IC can still induce cell death in a PKR independent manner (Figure 3A). We would like to suggest that if this strategy is translated successfully to the clinical setting it may indeed help patients with GBM. In fact, several neurosurgery teams are interested in implementing this therapy to clinic once all preclinical studies will be completed.
The elimination of EGFR over-expressing adenocarcinoma and breast cancer in vivo (Figure 7C) suggests that in principle this strategy can be applied to treat other cancers that over-express EGFR. It should be emphasized that there are a number of scenarios, other than GBM in which the local application of the EGFR targeted poly IC is advantageous, like head and neck cancer in which EGFR is over-expressed and local therapies are often used.
Finally, we would like to suggest that the strategy of ligand-guided delivery of dsRNA described here, can in principle, be applied to other cancers in which a particular receptor is over-expressed and undergoes endocytosis. Receptors that are over-expressed in many tumors and qualify as candidates for targeting poly(IC) include the transferrin receptor [26], the PDGF receptor [27] and the IGF-1 receptor [28].
Supporting Information
Figure S1 (poly IC)PEI-PEG-EGF(+PEI-Mel) Complexes Selectively Kill U87MGwtEGFR Cells
Cells were seeded onto a 96-well plate at a density of 5,000 cells in 0.2 ml of medium per well and grown overnight. Cells were then transfected as described [6,9] with poly IC at the indicated concentrations using either PEI-PEG-EGF or PEI-PEG-EGF+PEI-Mel (w/w ratio PEI-PEG-EGF:PEI-Mel = 1:10) complexes. The viability was measured by the CellTiter-Glo Luminescent Cell Viability Assay (Promega) according to the manufacturer's instructions at 1 h after transfection. Transfection with native PEI and Fugene6 (Roche) was used to control for non-specific effect of poly IC. HBS stands for transfection buffer alone [6].
(119 KB PDF).
Click here for additional data file.
Figure S2 Bystander Effect
For this assay, 500,000 U87MGwtEGFR cells were seeded onto 6-cm plates, grown overnight in 2 ml of medium, and transfected with poly IC at the indicated concentrations using PEI-PEG-EGF+PEI-Mel complexes. Medium was collected at the indicated times after transfection (x-axis). U87MG and U87MGΔEGFR “indicator” cells were seeded in 96-well plates (4,000 cells/well) and grown overnight in 200 μl of medium. 100 μl of medium was then replaced by the medium collected from the transfected U87MGwtEGFR cells. In NR samples, medium was not replaced. Growth inhibition was examined 48 h after medium exchange.
(35 KB PDF).
Click here for additional data file.
Patient Summary
Background
Glioblastomas are the most frequent and the most aggressive human brain tumors. Scientists have learned more about these tumors in the past few years, but there are still very few treatment options besides surgery and radiation, and only 3% of patients survive longer than 5 years after diagnosis.
Why Was This Study Done?
Over half of glioblastomas have a particular characteristic: the cells that make up the tumor have 10–100 times more of a particular molecule called EGFR on their surface than the surrounding non-tumor cells. Taking advantage of that fact, the researchers devised a strategy that used EGFR as a target to deliver a treatment that they hoped would kill the tumor cells.
What Did the Researchers Do and Find?
All of the experiments were done with cultured cells, and some of them used mice in which tumors had been implanted. The researchers created a mix of chemicals that would latch on to the EGFR on the tumor cell surfaces, be taken up by the tumor cells, and once inside the tumor cell would instruct it to commit suicide. They found that this approach worked in cultured tumor cells with lots of EGFR on their surfaces, and in mice that had brain tumors comprised of such cells. They found that the treatment was quite specific (that is, it killed the tumor cells and not the surrounding normal brain tissue) and effective (all of the tumor cells seemed to have been killed and the tumors didn't grow back).
What Does This Mean?
These results are encouraging. Many questions need to be answered before we know whether this approach or a similar one will be practical and efficient against human glioblastomas. Given the lack of current options, it certainly seems worth pursuing, and the researchers are currently preparing essential preclinical studies in which a clinical-grade drug will be prepared, and its toxicity, distribution to various organs, and other important analytical studies performed. At present the authors examine in more detail the toxicity profile of the poly IC complexed to the EGF receptor targeting vector.
Where Can I Find More Information Online?
The following Web sites provide information on brain cancer and glioblastomas.
National Brain Tumor Foundation in the US:
http://www.braintumor.org/
Cancer BACUP pages on brain cancer:
http://www.cancerbacup.org.uk/Cancertype/Brain
The Brain Tumor Society:
http://www.tbts.org/
NCI information pages on brain tumors:
http://www.cancer.gov/cancertopics/types/brain/
This study was supported by the Israel Cancer Association (Tel Aviv, Israel) and Algen Biopharmaceuticals (Jerusalem, Israel). We would like to thank Dr. Yael Friedman from our unit for the fluorescent labeling of poly IC. We also would like to thank Dr. Shoshana Klein from our unit for her comments and editing, and Wolfgang Rödl (from our unit at LMU) for the conjugate synthesis. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Citation: Shir A, Ogris M, Wagner E, Levitzki A (2006) EGF receptor-targeted synthetic double-stranded RNA eliminates glioblastoma, breast cancer, and adenocarcinoma tumors in mice. PLoS Med 3(1): e6.
Abbreviations
2-AP2-aminopurine
dsRNAdouble-stranded RNA
DTTdithiothreitol
EGFepidermal growth factor
EGFRepidermal growth factor receptor
GBMglioblastoma multiforme
mEGFrecombinant mouse EGF
MPPEmelittin-polyethyleneimine-polyethyleneglycol-EGF
NHS-PEG-MALN-hydroxysuccinimide-PEG maleimido
PEIpolyethylenimine
PEI25-PEG-EGFpolyethylenimine-polyethyleneglycol-EGF
PKRdouble-stranded RNA-dependent protein kinase
poly ICpolyinosine-cytosine
SCIDsevere combined immunodeficiency
==== Refs
References
Shir A Levitzki A Gene therapy for glioblastoma: Future perspective for delivery systems and molecular targets Cell Mol Neurobiol 2001 21 645 656 12043839
Liu TF Tatter SB Willingham MC Yang M Hu JJ Growth factor receptor expression varies among high-grade gliomas and normal brain: Epidermal growth factor receptor has excellent properties for interstitial fusion protein therapy Mol Cancer Ther 2003 2 783 787 12939468
Saunders LR Barber GN The dsRNA binding protein family: Critical roles, diverse cellular functions Faseb J 2003 17 961 983 12773480
Chawla-Sarkar M Lindner DJ Liu YF Williams BR Sen GC Apoptosis and interferons: Role of interferon-stimulated genes as mediators of apoptosis Apoptosis 2003 8 237 249 12766484
Parker LM Fierro-Monti I Reichman TW Gunnery S Mathews MB Double-stranded RNA-binding proteins and the control of protein synthesis and cell growth Cold Spring Harb Symp Quant Biol 2001 66 485 497 12762051
Ogris M Walker G Blessing T Kircheis R Wolschek M Tumour-targeted gene therapy: Strategies for the preparation of ligand-polyethylene glycol-polyethylenimine/DNA complexes J Control Release 2003 91 173 181 12932649
Wolschek MF Thallinger C Kursa M Rossler V Allen M Specific systemic nonviral gene delivery to human hepatocellular carcinoma xenografts in SCID mice Hepatology 2002 36 1106 1114 12395320
Shir A Levitzki A Inhibition of glioma growth by tumour-specific activation of double-stranded RNA-dependent protein kinase PKR Nat Biotechnol 2002 20 895 900 12205508
Ogris M Carlisle RC Bettinger T Seymour LW Melittin enables efficient vesicular escape and enhanced nuclear access of nonviral gene delivery vectors J Biol Chem 2001 276 47550 47555 11600500
Han Y Caday CG Nanda A Cavenee WK Huang HJ Tyrphostin AG 1478 preferentially inhibits human glioma cells expressing truncated rather than wild-type epidermal growth factor receptors Cancer Res 1996 56 3859 3861 8752145
Liu TF Cohen KA Ramage JG Willingham MC Thorburn AM A diphtheria toxin-epidermal growth factor fusion protein is cytotoxic to human glioblastoma multiforme cells Cancer Res 2003 63 1834 1837 12702570
Bettinger T Carlisle RC Read ML Ogris M Seymour LW Peptide-mediated RNA delivery: A novel approach for enhanced transfection of primary and post-mitotic cells Nucleic Acids Res 2001 29 3882 3891 11557821
Lamy L Ticchioni M Rouquette-Jazdanian AK Samson M Deckert M CD47 and the 19 kDa interacting protein-3 (BNIP3) in T cell apoptosis J Biol Chem 2003 278 23915 23921 12690108
Iordanov MS Ryabinina OP Schneider P Magun BE Two mechanisms of caspase 9 processing in double-stranded RNA- and virus-triggered apoptosis Apoptosis 2005 10 153 166 15711931
Ben-Asouli Y Banai Y Pel-Or Y Shir A Kaempfer R Human interferon-gamma mRNA autoregulates its translation through a pseudoknot that activates the interferon-inducible protein kinase PKR Cell 2002 108 221 232 11832212
Kershaw MH Wang G Westwood JA Pachynski RK Tiffany HL Redirecting migration of T cells to chemokine secreted from tumours by genetic modification with CXCR2 Hum Gene Ther 2002 13 1971 1980 12427307
Huang H Xiang J Synergistic effect of lymphotactin and interferon gamma-inducible protein-10 transgene expression in T-cell localization and adoptive T-cell therapy of tumours Int J Cancer 2004 109 817 825 15027114
Milas L Fan Z Andratschke NH Ang KK Epidermal growth factor receptor and tumour response to radiation: In vivo preclinical studies Int J Radiat Oncol Biol Phys 2004 58 966 971 14967457
Rae JM Lippman ME Evaluation of novel epidermal growth factor receptor tyrosine kinase inhibitors Breast Cancer Res Treat 2004 83 99 107 14997040
Comis RL The current situation: Erlotinib (tarceva(r)) and gefitinib (iressa(r)) in non-small cell lung cancer Oncologist 2005 10 467 470 16079313
Mendelsohn J Antibody-mediated EGF receptor blockade as an anticancer therapy: from the laboratory to the clinic Cancer Immunol Immunother 2003 52 342 346 12700950
Study JJ HER1/EGFR tyrosine kinase inhibitors for the treatment of glioblastoma multiforme J Neurooncol 2005 74 77 86 16078112
Byrne BJ Garst J Epidermal growth factor receptor inhibitors and their role in non-small-cell lung cancer Curr Oncol Rep 2005 7 241 247 15946581
Tsuchihashi Z Khambata-Ford S Hanna N Janne PA Responsiveness to cetuximab without mutations in EGFR N Engl J Med 2005 353 208 209 16014894
Friedrich I Eizenbach M Sajman J Ben-Bassat H Levitzki A A cellular screening assay to test the ability of PKR to induce cell death in mammalian cells Mol Ther 2005 12 969 975 16084774
Hogemann-Savellano D Bos E Blondet C Sato F Abe T The transferrin receptor: A potential molecular imaging marker for human cancer Neoplasia 2003 5 495 506 14965443
Varela M Ranuncolo SM Morand A Lastiri J De Kier Joffe EB EGF-R and PDGF-R, but not bcl-2, overexpression predict overall survival in patients with low-grade astrocytomas J Surg Oncol 2004 86 34 40 15048678
Moschos SJ Mantzoros CS The role of the IGF system in cancer: From basic to clinical studies and clinical applications Oncology 2002 63 317 332 12417786
|
16318410
|
PMC1298941
|
CC BY
|
2021-01-05 10:40:32
|
no
|
PLoS Med. 2006 Jan 6; 3(1):e6
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030006
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1631841410.1371/journal.pmed.0030007Research ArticleBioinformatics/Computational BiologyEcologyInfectious DiseasesEpidemiology/Public HealthHyperinfectivity: A Critical Element in the Ability of V. cholerae to Cause Epidemics? Epidemic CholeraHartley David M
1
*Morris J. Glenn Jr
1
Smith David L
2
1Department of Epidemiology and Preventive Medicine, University of Maryland School of Medicine, Baltimore, Maryland, United States of America2Fogarty International Center, National Institutes of Health, Bethesda, Maryland, United States of AmericaFerguson Neil Academic EditorImperial College LondonUnited Kingdom*To whom correspondence should be addressed. E-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
Author Contributions: DMH and DLS designed the study and analyzed the data. DMH, JGM, and DLS contributed to writing the paper.
1 2006 6 12 2005 3 1 e72 6 2005 30 9 2005 Copyright: © 2006 Hartley et al.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Adjusting Cholera Models to Recent Experimental Data
Background
Cholera is an ancient disease that continues to cause epidemic and pandemic disease despite ongoing efforts to limit its spread. Mathematical models provide one means of assessing the utility of various proposed interventions. However, cholera models that have been developed to date have had limitations, suggesting that there are basic elements of cholera transmission that we still do not understand.
Methods and Findings
Recent laboratory findings suggest that passage of Vibrio cholerae O1 Inaba El Tor through the gastrointestinal tract results in a short-lived, hyperinfectious state of the organism that decays in a matter of hours into a state of lower infectiousness. Incorporation of this hyperinfectious state into our disease model provides a much better fit with the observed epidemic pattern of cholera. These findings help to substantiate the clinical relevance of laboratory observations regarding the hyperinfectious state, and underscore the critical importance of human-to-human versus environment-to-human transmission in the generation of epidemic and pandemic disease.
Conclusions
To have maximal impact on limiting epidemic spread of cholera, interventions should be targeted toward minimizing risk of transmission of the short-lived, hyperinfectious form of toxigenic Vibrio cholerae. The possibility of comparable hyperinfectious states in other major epidemic diseases also needs to be evaluated and, as appropriate, incorporated into models of disease prevention.
Adjusting cholera models to take into account a short-lived hyperinfectious state proposed by recent experimental data improves their fit with real outbreak data.
==== Body
Introduction
While advances in medicine and public health have vanquished many pandemic diseases, 52 nations reported 142,311 cholera cases and 4,564 deaths in 2002, though these statistics are thought to represent a small subset of actual cases and deaths globally due to poor surveillance and under-reporting [1]. A complex web of interactions between the human host, pathogen, and environment are associated with the seasonal epidemics of cholera seen in endemic regions. The risk factors for cholera are varied and stem from multiple transmission pathways, including direct person-to-person contact and indirect transmission through the environment (e.g., food and water contamination) [2].
In epidemic situations, a fundamental question regarding the epidemiology of cholera is: what is the relative importance of human-to-human (i.e., fecal-oral) versus environment-to-human transmission (i.e., exposure to the environmental reservoir of Vibrio cholerae)? Answering this question may allow us to predict the onset, and potentially the intensity, of epidemics in endemic regions, as well as the speed and intensity of spread of cholera as it emerges in naïve regions. It may also afford us insight into new prevention, intervention, and control strategies to limit or prevent cholera transmission.
There are clues to the answer. Epidemic cholera is characteristically explosive in nature; when introduced into populations lacking prior immunity to the organism, spread through the population is measured in weeks, and involves all age groups [3,4]. Among more than 200 documented serogroups of V. cholerae, epidemic disease has been linked almost exclusively with serogroups O1 and O139. “Epidemic” strains colonize the small intestine and elaborate an enterotoxin (cholera toxin), which stimulates water and electrolyte secretion by intestinal endothelial cells and leads to massive fluid loss and profuse diarrhea. In volunteer studies, the frequency and severity of cholera has been correlated with inoculum [5,6]. The existence of such a dose-response relation (though typically difficult to measure) implies that the ability to infect a host is a key determinant of disease. Infective doses of environmental V. cholerae are thought to be in the range of 102–103 cells [2]. When individuals in a population are challenged with a dose many times larger than the ID50 (the infectious dose sufficient to produce frank disease in 50% of those exposed), the majority will be very likely to develop disease; an “explosive” epidemic will result. When challenged with a small fraction of the ID50, transmission will develop less rapidly, be less intense, and the outbreak will develop less rapidly. Thus, the inoculum of V. cholerae relative to the ID50 is key to understanding the intensity of cholera transmission.
In this context, recent experimental observations suggest that the V. cholerae ID50 depends upon the length of time the pathogen has existed outside the host. Passage of V. cholerae O1 Inaba El Tor through the human host appears to transiently increase the infectivity of V. cholerae [7]. Laboratory experiments demonstrate that when inoculated into the intestines of mice via gavage feeding, freshly shed V. cholerae greatly out-competes bacteria grown in vitro, by as much as 700-fold. The competitive advantage is short-lived; after standing 18 h, freshly shed V. cholerae organisms lose their competitive advantage. The advantage is maintained, however, for at least 5 h when incubated in pond water at room temperature. Comparing freshly shed vibrios to those not freshly shed, a different set of genes were up-regulated and these are thought to be responsible for faster bacterial growth in the gastrointestinal tract and increased shedding. Such observations suggest that passage of V. cholerae O1 Inaba El Tor through the human gastrointestinal tract results in a short-lived, hyperinfectious (HI) state. Recently, a second observation of a HI state in V. cholerae O1 El Tor has been published [8]. Whether similar states exist in other strains, biogroups, or serotypes of V. cholerae is unknown.
If indeed such a state exists, this hyperinfectivity is key to understanding the explosive nature of human-to-human transmission in outbreaks. Contact with freshly shed V. cholerae (lower ID50) is much more likely to cause disease than similar contact with V. cholerae shed much earlier (higher ID50). Moreover, since the decay from the HI state occurs within hours, it suggests that rapid local transmission through the direct route is responsible for rapid spread and the “explosive” nature of cholera epidemics, while the much slower environmental route accounts for much slower dynamics. Below we will use mathematical modeling to demonstrate that the existence of a transient HI state and the attendant reduction in ID50 explains the explosiveness of cholera epidemics. It is also consistent with other observed facets of cholera epidemiology and has concrete implications for public health strategies in prevention and intervention.
Methods
Codeço has modeled the traditional picture of cholera, in which the pathogen has no transient HI state (Figure 1A) [9]. The model applies to a population of N people who are born and die on average at the rate b. Infections, I, are caused by ingesting water contaminated with B vibrios per ml. Ingestion occurs at the rate β. When B equals κ, the probability of ingestion resulting in disease is 0.5. Cases contribute to V. cholerae in the aquatic environment at a rate ξ and cases cease to be infectious at the rate γ. Vibrios shed into the aquatic environment lose viability at the rate δ.
Figure 1 Models of Cholera Transmission
(A) Model of cholera transmission, neglecting a HI state of V. cholerae. Simple model reflecting a picture of contagion in which susceptibles (S) become infectious (I) after consuming concentrations (shown by B) of V. cholerae. Individuals may achieve a transient immunity (R) after recovery.
(B) Model of cholera transmission, incorporating a HI state of V. cholerae. This model includes contagion of V. cholerae that is HI (H) for a brief time after shedding and decays into a state of lower infectiousness (L). Greek letters denote rates of transitions between the different states, as described in Table 1.
Table 1 Model Parameters and Values
For any such model, the basic reproductive number, R0, provides a threshold for an epidemic to occur. Biologically speaking, R0 is the number of cases per case at the beginning of the outbreak. If R0 < 1, then a pathogen introduced into an immunologically naïve population will eventually die out. When R0 > 1, endemicity is possible. At the same time, R0 implicitly defines a timescale—the average time it takes for a pathogen to complete one generation. The larger R0 and the shorter the generation, the more explosive epidemic transmission will be. The basic reproductive number for the model in [9] (no HI state) is the product of three epidemiologic factors: the total amount of V. cholerae shed into the environment (the term in the first pair of parentheses in the expression for R0 below); the number of new cases per unit time generated by the shed pathogens (the term in the second pair of parentheses below); and the expected time that V. cholerae in the environment remains infectious before losing viability (the term in the third pair of parentheses below).
Estimation of the numerical value of R0 for any disease is a difficult task. The above expression is useful in that it illustrates the quantities upon which the basic reproduction number depends, and their relative importance. Estimating the absolute value of R0 from this expression, however, would entail knowledge of all variables in the expression, including some that are difficult to estimate directly. Many of these have been estimated in the literature (Table 1), but others are less well known and/or are highly variable. The parameter β (effectively, the population average rate of drinking potentially contaminated water) in particular is difficult to estimate. Codeço takes β = 1/day as a baseline, which yields R0 = 15. Alternatively, β can be estimated algebraically by approximating R0 roughly using the average age at first infection (A) and the mean lifetime of the population (L) in endemic areas.
This method for estimating R0 depends on many assumptions that may not be satisfied for cholera [10]. On the other hand, the L/A is a useful index of R0 for cholera in endemic areas since both L and A are readily measured. In four epidemics where estimates of L and A were both available, R0 values ranged from approximately 3 up to 15 (Table 2), with an average of ~ 8.7.
Table 2 Estimates of R0 from Past Cholera Epidemics
As described below, we have modified the model of cholera described in [9] to include a HI state of V. cholerae. The model is expressed in terms of differential equations and analyzed using standard methods [11,12]. The temporal epidemic and endemic behavior of cholera are estimated from the model through computer simulation of the model. The difference in the timing and intensity of epidemics that the HI state introduces relative to a single state of lower infectivity is illustrated by plotting simulated cholera cases through time when the HI state is both included in and excluded from the model. All simulations are carried out using the R (version 2.1.1, http://www.r-project.org/) odesolve library.
We have extended the Codeço model to incorporate a state of hyperinfectivity (Figure 1B). In this modified model, infections, I, are caused by ingesting water contaminated with BH HI vibrios per ml or BL non-HI vibrios per ml. Ingestion of HI vibrios occurs at the rate βH while ingestion of non-HI vibrios occurs at the rate βL. When BH equals κH, the probability of ingestion resulting in disease is 0.5, and similarly for BL and κL. In other words, the models assume that the relationship between infection rates and the density of cholera is described by a saturating function β B/(B + κ). Vibrios in the HI state decay into a state of lower infectivity (“non-HI”) at the rate χ. Cases shed HI V. cholerae into the aquatic environment at a rate ξ and cases cease to be infectious at the rate γ. Non-HI vibrios shed into the aquatic environment lose viability at the rate δL.
These ideas are expressed in terms of the following set of differential equations:
Representative values of the parameters appearing in these equations are listed in Table 1.
Results
The expression for the basic reproductive number for this model is more complicated than in the simpler model and involves contributions from both the HI and non-HI states. By computing the dominant eigenvalue of the next-generation matrix corresponding to these equations, using the method described in Watmough and van den Driessche [12], it follows that
This is a new result. Biologically speaking, the first term in the parenthesis is associated with the number of new infections caused by HI vibrios, and the second term is associated with new infections caused by non-HI V. cholerae. As before, ξ/(γ + μ) is the average amount of V. cholerae shed per individual. The terms 1/χ and 1/δL are the expected times that vibrios remain in the HI and non-HI infectious states, respectively, before they decay into a non-infectious state (i.e., die in the environment or lose viability). Finally, βH /κH and βL /κL are the number of new cases generated in terms of the HI and non-HI vibrios, respectively, per unit time as measured by the number of ID50 concentrations.
An important question in cholera epidemiology is the relative importance of the HI and non-HI infectious routes. This quantity probably varies from place to place, depending on sanitation, population density, and hygiene standards. However, several relevant points can be made under a restricted set of assumptions.
First, we consider the case where contact with HI vibrios is about as frequent as contact with non-HI vibrios (βL ~ βH). In such cases, the ratio of the HI to non-HI contributions to the R0 is approximately (κLδL)/(κHχ). Using parameter values cited in Table 1, this quantity is approximately 4.7, suggesting that the HI state is about five times more important than the non-HI state in generating cases early in the epidemic. As a thought experiment, consider a severe epidemic with both modes of transmission (R0 ~ 18.2). If all the transmission from HI cholera were somehow prevented, but environmental (i.e., non-HI) transmission continued, the spread would be severely reduced (to R0 ~ 3.2). In other words, the number of cases per case would be reduced by a factor of 5.7 at the beginning of the epidemic if there were no HI state.
We have used the model and the parameters in Table 1 to simulate the epidemic curves that might be observed in hypothetical community (N = 10,000 individuals; average lifespan = 1/b = 30 y) in the case that (a) no HI transmission exists versus (b) the case when a HI transmission occurs. We assume that a cholera epidemic begins when a single infectious individual enters a completely susceptible population and that the aquatic reservoir initially contains no V. cholerae. Figure 2 shows the number of cases as a function of time for the two scenarios. With no HI state (Figure 2A), cases peak after about 25 wk, at which time approximately 500 individuals are infectious, then decline until essentially none are present 50 wk following the introduction. With the HI state, (Figure 2B), cases peak in less than 2 wk, at which time approximately 3,500 people are infectious, then decline rapidly approximately 6 wk after the introduction. While, in reality, the depletion of susceptible population in the local neighborhood, control measures, and other changes in human behavior would slow the spread of cholera, the model qualitatively illustrates the role of the HI state in the timing and intensity of cholera outbreaks in naïve populations.
Figure 2 Number of Cholera Cases as a Function of Time
(A) Model results when there is no HI state of V. cholerae. (B) Model results when there is a HI state of V. cholerae. Initial conditions for both (A) and (B): No V. cholerae present in the environment and all persons are susceptible except for a single active case present at the onset of the simulation.
Second, we make the stronger assumption that HI and non-HI cholera make equal contributions to R0; hence, in this scenario, there is less contact that occurs with HI cholera compared with non-HI cholera. Despite the reduction in the relative contact rates with HI cholera so that its contribution is equal to that of the non-HI, transmission through the HI route generates a greater share of the cases because it is able to complete several transmission cycles by the time one complete cycle is completed through the non-HI route. In other words, the generation time is substantially shorter when transmission occurs through direct transmission of HI cholera, and thus, cholera epidemics tend to be more explosive.
To illustrate, we set the contact rates so that the cases per case are equal (this occurs when βH = βL κH χ/(κLδL) and then plot the incidence (new infections per unit time) that occur from each transmission mode:
Incident cases due to the HI state correspond to j = H (λH; broken trace in Figure 3A) and incident cases due to the non-HI state correspond to j = L (λL; solid trace). Note that HI transmission produces the majority of new infections: in this case, the ratio of the height of the peaks is approximately 1.6. Moreover, the incidence due to the HI state peaks approximately 2 wk before the non-HI incidence peaks, and decays more rapidly than the incidence due to the non-HI state. The non-HI state is responsible for the slower dynamics of the outbreak while the HI state drives the fast dynamics. Another illustration of this point is depicted in Figure 3B, where we plot the relative contribution of each state to incident cases, λH/(λH + λL) (broken trace) and λL/(λH + λL) (solid trace) throughout the outbreak. From the beginning of the outbreak through approximately 5 wk, the HI state causes more cases than does the non-HI state. Thereafter, when most of the susceptible population has been infected, the non-HI state becomes more important.
Figure 3 Roles Played by HI and Non-HI V. cholerae through the Course of an Epidemic when Each State Contributes Equally to R0
Broken traces correspond to the HI state, and solid traces correspond to the non-HI state. Initial conditions are the same as in Figure 2B. (A) Rates at which incident infections appear. (B) Relative contribution of each state to incident cases.
The analysis above suggests that the HI state is extremely important in the transmission of epidemic cholera. How sensitive are these results to the particular model parameters used? Of all the parameters, βH and βL are most important for control and least well known. In order to illustrate the sensitivity of the model to the relative variation of these parameters, in Figure 4 we plot how R0 depends upon the relative rates of contact with contaminated water, βL/βH, when the parameters of Table 1 are used. When βL ~ βH, (contact with water contaminated with HI vibrios is as frequent as contact with water containing non-HI vibrios) R0 is approximately 18.2, and as βH becomes much smaller than βL (contact with water contaminated with HI vibrios is much less frequent then contact with water containing non-HI vibrios), R0 approaches 3.2, the value we expect when there is no contact with HI V. cholerae. When contact with water contaminated with HI vibrios is much more frequent then contact with water containing non-HI vibrios, R0 will be very large (> 18.2).
Figure 4 Sensitivity of the Basic Reproduction Ratio, R0, to the Relative Contact Rate with HI and Non-HI V. cholerae (βL/βH)
Discussion
The epidemiological significance of the HI state is 2-fold. First, because HI vibrios are recently shed from individuals, they are close to humans, compared with vibrios in the environment. Consequently, HI vibrios are more likely to come into contact with other individuals (βH > βL in the model). The rapid decay of HI V. cholerae also implies that new infections will tend to be associated with recent infection of an infectious individual. Thus, even if HI vibrios are transmitted indirectly through local contamination (e.g., water or food), transmission will behave dynamically in populations as if it were direct. Second, the dramatic competitive advantage observed in [7] suggests the ID50 for HI V. cholerae is much lower than for non-HI V. cholerae (and thus κH << κL in the model). Even if the rates of exposure from quasi-direct transmission and indirect transmission through the environment are the same, for example, in the case of drinking well-mixed water from a contaminated supply such as a river polluted with raw sewage, V. cholerae in the HI state may be more epidemiologically significant because it requires a lower dose to cause disease and because the generation times are shorter. As demonstrated above, when βL ~ βH, for the parameters estimated from the literature (Table 1), V. cholera in the HI state contributes nearly six times more secondary cases than vibrios in the non-HI state at the beginning of the epidemic. Thus, we propose that these two effects combine to generate the explosive epidemic behavior historically observed in cholera outbreaks.
From a public health vantage point, these results suggest that careful attention should be given to differentiating the risk of exposure to “fresh” fecal contamination, versus “older” organisms acquired from environmental sources. Freshly excreted organisms, being many more times more infectious than less-recently excreted vibrios, are expected to dominate cholera transmission when proper hygiene is not practiced and effective waste disposal is not available. Likewise, older organisms are expected to dominate cholera transmission dynamics when good hygiene and effective waste engineering practices are followed. Real epidemics may combine elements of both—isolated cases from environmental cholera followed by clusters of cases due to local transmission. Such results re-emphasize the importance of focusing prevention efforts on improving hygiene, including frequent hand-washing, sewage disposal, good water, and food-handling practices in the home and in food service establishments (restaurants, street vendors) where contact with freshly shed vibrios is most likely.
Interestingly, these results suggest a possible explanation for the puzzle of why cholera was controlled in Europe and the United States before the widespread disinfection of water. In the mid- to late-1800s, sewage disposal systems began to be engineered and built in municipal areas. Because these systems dramatically increased hygiene standards and decreased the opportunity for contact with fresh feces (resulting in a smaller βH), the HI state of the pathogen played a less important a role in transmission. Of course, if such systems transported sewage to locations where non-HI vibrios could contaminate drinking sources, transmission remained possible, albeit at a greatly reduced rate. We postulate that the HI state would have decayed during transport, so that disease due to contaminated water supply would have been associated with non-HI V. cholerae. In fact, in the US decades passed before cholera was effectively wiped out following introduction in the early 1830s [13]. This provides an important implication for developing regions vulnerable to cholera outbreaks: even if water purification is not possible, there are still public health benefits to improved hygiene and sewage management systems.
The observation of the HI state was made experimentally for V. cholerae O1 Inaba El Tor [7]. To our knowledge, no observations under field conditions have appeared in the literature. Additional research is therefore needed to further characterize such states. The recent description of hyperinfectious V. cholerae O1 El Tor in a mouse model should aid in such research [8]. Our mathematical model demonstrates the potential importance of such a transient state for epidemic transmission, and highlights the need to search for analogous states in related species of V. cholerae. Moreover, if transient states of elevated infectivity exist in other pathogens besides V. cholerae, we anticipate they will be if key importance to epidemic transmission.
Patient Summary
Background
Cholera remains a public health problem in countries without access to safe drinking water and adequate sanitation. People can get cholera from contaminated food or water, or from contact with feces or vomit of patients. Researchers are trying to develop theoretical models for infectious diseases, including cholera, that allow them to understand how an outbreak happens, how it could best be contained, and how it might be prevented. Existing models have not been able to explain actual cholera outbreaks very well.
Why Was This Study Done?
Three years ago, another group of researchers reported that cholera bacteria isolated from the stools of sick patients were much more infectious than those found in contaminated water. (They compared the two by exposing mice to a mix and determining which bacteria made the mice sick.) Those researchers proposed that the infection of a human patient (i.e., the exposure to an environment that is quite different from their regular freshwater ponds) changes the cholera bacteria. As a result, for a short period of time, the bacteria become more infectious. The researchers who did the present work are interested in cholera modeling. They wanted to incorporate the idea of a short-lived hyperinfectious state and see whether the resulting model might better explain cholera outbreaks.
What Did the Researchers Do and Find?
They assumed that the number of hyperinfectious bacteria (i.e., those from fresh feces of a sick patient) necessary to cause disease was smaller than the number from contaminated food or water and that this period of hyperinfectivity was very short. They adjusted a model to take this into account and found that it could now better explain the explosive nature of cholera outbreaks.
What Does This Mean?
The original results suggesting that passage through humans makes cholera bacteria more infectious were based on experiments in mice. That a cholera model fits reality better when it assumes the existence of such a hyperinfectious state makes it likely that the findings are relevant to disease transmission in humans. The findings also have public health implications: they emphasize the need to avoid contamination with fresh patient feces in the context of overall improved hygiene.
Where Can I Find More Information Online?
The following Web sites provide information about cholera.
Wikipedia pages on cholera:
http://en.wikipedia.org/wiki/Cholera
WHO pages on cholera:
http://www.who.int/topics/cholera/en/
Cholera pages from the U.S. Centers for Disease Control and Prevention:
http://www.cdc.gov/ncidod/dbmd/diseaseinfo/cholera_g.htm
DMH is supposted by NIH Career Development Award (K25) AI-58956. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Citation: Hartley DM, Morris JG Jr, Smith DL (2006) Hyperinfectivity: A critical element in the ability of V. cholerae to cause epidemics? PLoS Med 3(1): e7.
Abbreviation
HIhyperinfectious
==== Refs
References
World Health Organization Cholera, 2002 Wkly Epidemiol Rec 2003 31 269 276
Kaper JB Morris JG Levine MM Cholera Clin Micro Rev 1995 8 48 86 7704895
Morris JG
Vibrio cholerae O139 Bengal: Emergence of a new epidemic strain of cholera Infect Agents Dis 1995 4 41 46 7728355
Koo D Traverso H Libel M Drasbek C Tauxe R Epidemic cholera in Latin America, 1991–1993: Implications of case definitions used for public health surveillance Bull Pan Am Health Organ 1996 30 134 143 8704754
Cash RA Music SI Libonati JP Synder MJ Wenzel RP Response of man to infection with Vibrio cholerae . I. Clinical, serologic, and bacteriologic responses to a known inoculum J Infect Dis 1974 129 45 52 4809112
Levine MM Nalin DR Rennels MB Hornick RB Sotman S Genetic susceptibility to cholera Ann Hum Biol 1979 6 369 374 394667
Merrell DS Butler SM Qadri F Dolganov NA Alam A Host-induced epidemic spread of the cholera bacterium Nature 2002 417 642 645 12050664
Alam A Larocque RC Harris JB Vanderspurt C Ryan ET Hyperinfectivity of human-passaged Vibrio cholerae can be modeled by growth in the infant mouse Infect Immun 2005 73 6674 6679 16177344
Codeço CT Endemic and epidemic dynamics of cholera: The role of the aquatic reservoir BMC Infect Dis 2001 1 1 11208258
Heffernan JM Smith RJ Wahl LM Perspectives on the basic reproductive ratio J Roy Soc Lond Interface 2005 2 281 293
Hethcote HW The mathematics of infectious diseases SIAM Rev 2000 42 599 653
van den Driessche P Watmough J Reproduction numbers and sub-threshold endemic equilibria for compartmental models of disease transmission Math Biosci 2002 180 29 48 12387915
Rosenberg CE The cholera years: The United States in 1832, 1849, and 1866 1962 Chicago The University of Chicago Press
Tudor V Strati I Smallpox, cholera 1977 Tunbridge Wells (England) Abacus Press 313
Hendrix TR The pathophysiology of cholera Bull N Y Acad Med 1971 47 1169 1180 4329549
Siddique AK Zaman K Baqui AH Akram K Mutsuddy P Cholera epidemics in Bangladesh: 1985–1991 J Diarrhoeal Dis Res 1992 10 79 86 1500643
Simanjuntak CH Larasati W Arjoso S Putri M Lesmana M Cholera in Indonesia in 1993–1999 Am J Trop Med Hyg 2001 65 788 797 11791976
Nizami SQ Farooqui BJ Cholera in children in Karachi from 1990 through 1995: A study of cases admitted to a tertiary care hospital J Pak Med Assoc 1998 48 171 173 9813989
Sack RB Siddique AK Longini IM Nizam A Yunus M A 4-year study of the epidemiology of Vibrio cholerae in four rural areas of Bangladesh J Infect Dis 2003 187 96 101 12508151
|
16318414
|
PMC1298942
|
CC BY
|
2021-01-05 10:40:32
|
no
|
PLoS Med. 2006 Jan 6; 3(1):e7
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030007
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1631841510.1371/journal.pmed.0030013Research ArticleBioinformatics/Computational BiologyBiotechnologyCancer BiologyGenetics/Genomics/Gene TherapyMedical InformaticsNephrologyOncologyPathologyUrologyGeneticsMedical InformaticsOncologyPathologyRenal MedicineGene Expression Profiling Predicts Survival in Conventional Renal Cell Carcinoma Gene Expression and Survival in Renal CancerZhao Hongjuan
1
Ljungberg Börje
2
Grankvist Kjell
2
Rasmuson Torgny
2
Tibshirani Robert
3
Brooks James D
1
*1Department of Urology, Stanford University School of Medicine, Stanford, California, United States of America2Departments of Surgical and Perioperative Sciences, Urology, and Andrology, Medical Biosciences, Clinical Chemistry, and Radiation Sciences, Oncology, Umeȧ University, Umeȧ, Sweden3Department of Health Research and Policy, Stanford University School of Medicine, Stanford, California, United States of AmericaMarincola Francesco Academic EditorNational Institutes of HealthUnited States of America*To whom correspondence should be addressed. E-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
Author Contributions: BL and JDB designed the study. HZ, BL, KG, and TR performed the experiments. HZ, RT, and JDB analyzed the data. HZ, BL, RT, and JDB contributed to writing the paper.
1 2006 6 12 2005 3 1 e1318 7 2005 12 10 2005 Copyright: © 2006 Zhao et al.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Expression Profiling Predicts Survival in Kidney Cancer
Background
Conventional renal cell carcinoma (cRCC) accounts for most of the deaths due to kidney cancer. Tumor stage, grade, and patient performance status are used currently to predict survival after surgery. Our goal was to identify gene expression features, using comprehensive gene expression profiling, that correlate with survival.
Methods and Findings
Gene expression profiles were determined in 177 primary cRCCs using DNA microarrays. Unsupervised hierarchical clustering analysis segregated cRCC into five gene expression subgroups. Expression subgroup was correlated with survival in long-term follow-up and was independent of grade, stage, and performance status. The tumors were then divided evenly into training and test sets that were balanced for grade, stage, performance status, and length of follow-up. A semisupervised learning algorithm (supervised principal components analysis) was applied to identify transcripts whose expression was associated with survival in the training set, and the performance of this gene expression-based survival predictor was assessed using the test set. With this method, we identified 259 genes that accurately predicted disease-specific survival among patients in the independent validation group (p < 0.001). In multivariate analysis, the gene expression predictor was a strong predictor of survival independent of tumor stage, grade, and performance status (p < 0.001).
Conclusions
cRCC displays molecular heterogeneity and can be separated into gene expression subgroups that correlate with survival after surgery. We have identified a set of 259 genes that predict survival after surgery independent of clinical prognostic factors.
Molecular heterogeneity of renal cell carcinomas can be used to distinguish subgroups that correlated with long term survival.
==== Body
Introduction
Nearly half of the patients diagnosed with renal cell carcinoma (RCC) succumb to their disease, and RCC accounts for 95,000 deaths per year worldwide [1]. In the United States, approximately 36,160 cases will be diagnosed this year alone, and 12,660 patients will die of their disease [2]. Conventional renal cell carcinoma (cRCC) accounts for approximately 75% of all RCC and accounts for the majority of kidney cancer mortality. Surgery (nephrectomy) can cure 60%–70% of patients with localized disease and prolong survival in patients with metastatic disease, although survival rates after treatment have not changed appreciably in the past 30 y [2,3]. Cytokine therapy, which is reserved for patients with advanced disease, can produce partial responses in 10%–15% of patients and durable remissions in 5% [4].
Tumor stage is the most powerful predictor of outcome in patients with cRCC, although it provides a relatively crude estimate of survival that limits its use in clinical decision making [5]. Several prognostic algorithms have been developed that incorporate tumor stage, grade, and patient performance status, and they predict survival better than stage alone [5–7]. Based on these algorithms, fewer radiographic imaging and blood tests have been proposed for patients predicted to have a low risk of recurrence after surgery, and adjuvant therapy has been suggested for high-risk patients. Unfortunately, many patients fall into intermediate-risk categories, and these algorithms do not predict survival or response to therapy in patients with advanced disease [6].
The limitations of the prognostic algorithms and the varied response to surgery and immunotherapy suggest that cRCCs are molecularly diverse and that capturing relevant molecular features could improve outcome prediction. In support of this idea, several small series used DNA microarray analysis to identify genes whose expression levels correlated with survival in RCC, although the prognostic gene sets did not overlap, and neither study has been validated independently [8–10]. To identify gene expression correlates of survival in cRCC, we used DNA microarrays to explore systematically the molecular variations underlying the biologic and clinical heterogeneity in a set of 177 tumors with associated detailed clinical information, including long-term follow-up.
Methods
Samples
Tumors from 177 consecutive patients who underwent radical nephrectomy for cRCC collected between 1985 and 2003 were selected from the fresh-frozen tissue bank in the Department of Urology, Umeȧ University Hospital (Umeȧ, Sweden). Written informed consent was obtained from all patients, and the study was approved by the institutional review board of each participating center. Patients in the study included 102 men and 75 women with cRCC diagnosed on the nephrectomy specimens by pathologists at Umeȧ University Hospital (summarized in Table 1). Mean age of the patients was 65 y (range, 34 to 85 y), and performance status, assessed using World Health Organization criteria, ranged from 0 (65 patients), 1 (64 patients), 2 (37 patients), 3 (ten patients), to 4 (one patient). Pathologic stage grouping of patients in the study, based on preoperative radiographic studies and pathological assessment of the surgical specimens was I (49 patients), II (29 patients), III (40 patients), and IV (59 patients). No patient received neoadjuvant therapy prior to surgery. Adjuvant interferon therapy was given to seven patients and adjuvant hormonal therapy to 12 patients, and all had stage IV disease at the time of surgery. Thirteen patients who recurred after surgery received salvage interferon therapy, nine had resection of metastases, and 19 received hormonal therapy. Patient follow-up status was assessed at least yearly by routine clinical follow-up at Umeȧ University Hospital or by contacting patients directly. Median follow-up of censored patients was 76 mo (range 19 to 224 mo). During the follow-up period, 87 patients died of their disease, 25 died of other causes, nine were alive with disease, and 56 were alive and free of disease.
Table 1 Description of Patients
Gene Expression Profiling
Total RNA was isolated from the cRCC tissue samples using TRIzol reagent (Invitrogen, Carlsbad, California, United States), according to the manufacturer's recommendations. The integrity of the total RNA was assessed using a 2100 Bioanalyzer (Agilent Technologies, Palo Alto, California, United States). Cy5-labeled total RNA from cRCC samples was mixed with Cy3-labeled Universal Human Reference RNA (Stratagene, La Jolla, California, United States) and hybridized to cDNA microarrays (manufactured by the Stanford Functional Genomics Facility) that contained over 40,000 cDNA clones, representing 27,290 unique UniGene clusters as described previously [11]. Arrays from ten different print runs were used in the study, and all arrays passed a set of quality control criteria defined by GenePix software, including mean of median background less than 500, feature variation less than 0.5, background variation less than 0.5, and features with saturated pixels less than 0.1%. For an explanation of each of these quality control measures, see http://www.moleculardevices.com/pages/software/gn_genepix_pro.html. Microarrays were imaged using an Axon GenePix 4000B scanner (Axon Instruments, Union City, California, United States), and fluorescence ratios of the tumor RNA specimens compared to the reference RNA were determined using GenePix software. Data was entered into Stanford Microarray Database for subsequent analysis [12]. The complete microarray dataset is available at http://smd.stanford.edu/cgi-bin/publication/viewPublication.pl?pub_no=484. The data also have been deposited in National Center for Biotechnology Information's Gene Expression Omnibus (see Accession Numbers section).
Statistical Analysis
Hierarchical clustering analysis
Fluorescence ratios were normalized by mean-centering genes for each microarray, mean-centering each gene across all microarrays, and centering within each of ten microarray print runs (to minimize potential print-run-specific bias). We selected 3,674 genes represented by 5,560 clones on the microarrays whose expression was both well measured and highly variable among samples (a complete list is available at http://smd.stanford.edu/cgi-bin/publication/viewPublication.pl?pub_no=484). We defined well-measured genes as those with a ratio of signal intensity to background noise of more than 1.5 for either the Cy5-labeled cRCC sample or the Cy3-labeled reference sample, in at least 70% of the samples hybridized. Genes with highly variable expression were defined as those whose expression was higher or lower by a factor of at least three than the average expression of all cRCC samples in at least ten cRCC samples. We applied two-way (genes-against-samples) average-linkage hierarchical clustering and used TreeView to visualize the results [13]. We compared the survival times of the five gene expression subgroups using Kaplan-Meier survival analysis and the log-rank test.
Supervised principal components analysis
For outcome prediction, we randomly divided samples—that had been prestratified to ensure that a similar proportion of samples in each group were from patients who had died and with similar clinical parameters, including tumor stage grouping, grade, performance status, and length of follow-up—into a separate training set (88 samples) and test set (89 samples) (Table 2). In the training set, we calculated modified univariate Cox proportional-hazard scores for all genes (n = 14,814) that were well measured to identify genes whose expression correlated with the duration of survival. (The modification adds a constant to the denominator, as described in [14].) We selected a set of genes whose absolute Cox score statistic exceeded a threshold that was chosen using multiple 2-fold cross-validation. To determine the threshold, the training samples were divided randomly and principal components were derived from half of the samples and then used in a Cox model to predict survival in the other half. We repeated this entire process five times and found that a threshold of ±1.5 yielded the highest average partial log-likelihood ratio statistic. Principal components analysis was then performed on all cases in the training set, using 340 transcripts representing 259 genes whose absolute Cox score equaled or exceeded the threshold. Only the first principal component was associated significantly with survival. For patients in the test set, a continuous risk score (that is, the supervised principal components [SPC] risk score) was calculated for each patient, based on transcript levels across the 340 transcripts and the weights assigned to each transcript derived from SPC analysis of the training set. Multivariate proportional-hazards analysis was performed on the test set with the SPC risk score as a continuous variable, along with stage grouping, grade, performance status, and gene expression subgroup derived from hierarchical clustering analysis.
Table 2 Patient Distribution between Training and Test Set
To evaluate the gene set as a categorical predictor of survival, we divided the training set into tertiles based on the SPC risk scores. The test set then was divided into three groups based on the tertiles of the SPC risk scores of the training set. We compared the survival times of the three subgroups in the training and test sets using Kaplan-Meier survival analysis and the log-rank test.
SPC analysis and multivariate proportional-hazards analysis were performed with the use of the R software package (available at www.r-project.org) and the superpc R package (available at http://www-stat.stanford.edu/~tibs/superpc). Kaplan-Meier survival analysis was performed with WinStat software (R. Fitch Software, Staufen, Germany).
Results
Gene Expression Profiles of cRCCs
Hierarchical clustering analysis of the 177 patient samples described in Table 1 was performed using 5,560 clones representing 3,674 unique genes whose expression varied more than 3-fold from the mean expression ratio (specimen RNA/reference RNA) in at least ten samples (Figure 1). A detailed view of the sample cluster dentodrogram is displayed in Figure S1. Tumors were partitioned into two main groups and five subgroups based on the differential expression of these 3,674 genes (Figure 1A). The grouping of the tumors in the dendrogram did not appear to be an artifact of the genes used to generate the cluster because varying data-filtering criteria (and the number of genes used in the hierarchical clustering analysis) resulted in a similar pattern of specimen clustering. A large and diverse set of genes distinguished the two main groups of tumors, all of which showed relatively high expression in tumors in subgroups 1 and 2 compared to subgroups 3, 4, and 5 (black bar in Figure 1A). These genes are involved in a variety of biological processes, including angiogenesis (FLT1, EPAS1, and JAG1), the Wnt signaling pathway (FZD1, FZD4, and TCF4), cell adhesion (CDH13, PECAM1, and VCAM1), and cellular metabolism (UGT2B7, UGT2B4, and GSTA2).
Figure 1 Unsupervised Hierarchical Clustering Analysis of 177 cRCCs
(A) Overview of the gene expression patterns of 3,674 genes whose expression varied more than 3-fold in at least ten samples across the 177 samples. Each row represents a single gene, and each column an experimental sample. Colored bars identify the locations of the inserts in (C–J). The degree of color saturation corresponds with the ratio of gene expression shown at the top of the image.
(B) Dendrogram representing similarities in the expression patterns between experimental samples. Samples were separated into two main groups and five subgroups (one in purple, two in blue, three in dark green, four in orange, and five in light blue) by the clustering algorithm.
(C) Hypoxia-induced gene cluster.
(D) Collagen gene cluster.
(E) Proliferation gene cluster.
(F–I) Genes distinguishing the two main groups (subgroups 1 and 2 from subgroups 3, 4, and 5).
(H) Energy generation gene cluster.
(J) Genes downregulated uniquely in subgroup 5.
Each of the five subgroups of tumors displayed distinct gene expression patterns. Examples of clusters whose expression patterns distinguished between the subgroups are shown in Figure 1B. Expression patterns in subgroups 1 and 2 were largely similar, although they differed in a set of genes involved in diverse biological processes, including the transcriptional regulators MLL3, EYA3, JMJD1C, CNOT4, CNOT6L, SP3, and TEAD1 (Figure 1I). Compared to the other cRCCs, those in subgroup 4 showed lower expression of many hypoxia-regulated genes (e.g., HIG2, EGLN3, CA9, and STC2) (Figure 1C). Conventional RCCs commonly harbor VHL gene mutations that result in increased expression of hypoxia-regulated genes, suggesting that subgroup 4 cancers either lack inactivating VHL mutations or downregulate hypoxia signaling pathways [15,16]. Subgroup 4 tumors also showed increased expression of many genes that characterize chromophobe carcinomas and oncocytomas, including KIT and the mitochondrial genes NNT, FH, GOT1, GOT2, SLC25A5, ATP2B1, ATP5G3, ATP5B, and ATP6V1A (Figure 1H). We have previously observed similar expression patterns in a subset of cRCCs that have granular cytoplasm [17], and a review of the pathological specimens revealed that 11 of 13 tumors in subgroup 4 were conventional carcinomas with granular cytoplasm. Subgroup 3 showed much higher expression of proliferation-associated genes compared to other tumors (CDCA3, CDC2, CENPE, CENPF, RRM2, and CCNB2), suggesting a higher proliferative activity in these tumors [18,19] (Figure 1E). Interestingly, there was little correlation between expression levels of the hypoxia-regulated genes and proliferation-associated genes (Pearson's correlation coefficient of 0.22), suggesting that higher proliferation activity does not render cRCCs hypoxic and highlighting that expression of hypoxia-regulated genes is an intrinsic feature of most cRCCs. Subgroup 3 tumors (and some subgroup 5 tumors) also showed high expression of several collagen genes (COL12A1, COL3A1, COL6A1, COL1A1, and COL5A2), and high expression of collagen genes has been associated with poor prognosis in several tumor types [20] (Figure 1D). Subgroup 5 uniquely displayed decreased expression of a large set of genes that prominently included several membrane transporters (NUP54, VPS54, STAM2, MAPK8IP3, G3BP2, and SLC30A9) (Figure 1J). The distinct gene expression profiles of each of the subgroups suggest that cRCCs are molecularly heterogeneous despite their similar histological appearance.
Gene Expression Subgroups of cRCC Differ in Their Clinical Behavior
The gene expression subgroups did not simply reflect differences in stage, grade, or performance status since none of these clinical parameters was significantly associated with tumor subtype (p > 0.5 by the chi-square test) (Figure 2A and 2B). The two main groups of cRCC defined by unsupervised hierarchical clustering analysis showed a small but significant difference in survival (subgroups 1 and 2 compared to subgroups 3, 4, and 5, p = 0.002 by the log-rank test) (Figure 2C). The five expression subgroups better defined classes of tumors that differed in their long-term survival. Kaplan-Meier analysis showed that patients with tumors in subgroup 3 had the worst outcome and those in subgroups 1 and 2 the best compared to other subgroups (p < 0.001 by the log-rank test) (Figure 2D). Furthermore, multivariate analysis showed that the expression subgroup was a powerful predictor of survival and was independent of grade, stage grouping, and performance status (p = 0.005, by the Cox model likelihood ratio test). Therefore, gene expression profiles separate cRCC into five subgroups that differ prognostically and that reflect differences in the behavior of the tumors not captured by stage, grade, and performance status.
Figure 2 Relationship of Gene Expression Subgroups to Clinical Parameters and SPC Risk Score in 177 cRCCs
(A) Dendrogram from the hierarchical cluster, with the clinical information for each of the samples. Subgroups are color-coded as in Figure 1. Color shade corresponds to the ranges of each of the clinical parameters displayed. Expected survival times for the censored observations were estimated from the Kaplan-Meier curve for all patients.
(B) Distribution of stage, grade, and patient performance status among five subgroups.
(C) Kaplan-Meier estimates of disease-specific survival in the two main gene expression groups of patients (subgroups 1 and 2 shown by the red bar below the dendrogram, compared to subgroups 3, 4, and 5 designated by the green bar).
(D) Kaplan-Meier estimates of disease-specific survival in the five subgroups of patients.
The X symbols in (C) and (D) denote censored data.
Gene Expression-Based Survival Predictor
Having identified gene expression signatures in tumors at the time of diagnosis that predict outcome by unsupervised methods (i.e., based purely on gene expression signatures intrinsic to the tumors), we attempted to define a gene expression-based survival predictor by correlating survival time with the gene expression signatures. We have found that supervised analyses that correlate gene expression with disease recurrence or survival (as binary outcome variables) or duration of survival are overly simplistic models of these complex datasets and in general not very accurate at predicting clinical outcomes. We have instead used “semisupervised” learning approaches to identify gene sets associated with survival in adult acute myeloid leukemia and diffuse B-cell lymphoma and have shown that they better identify gene expression signatures that are correlated with outcome compared to unsupervised and supervised methods of data analysis [11,21]. Whereas unsupervised methods assign tumors to a class based solely on gene expression, and supervised approaches use clinical outcome data to select genes associated with prognosis, semisupervised methods combine the advantages of both.
To identify genes highly correlated with survival in cRCC, we used SPC analysis, a novel semisupervised approach we have developed recently [14,22]. Patient samples were divided randomly into a training set of 88 cases and a test set of 89 cases that were balanced for stage grouping, grade, patient performance status, and length of follow-up (Table 2). In the training set, a modified Cox score was calculated for all well-measured genes, and genes whose Cox score exceeded a threshold that best predicted survival were used to carry out unsupervised principal components analysis (Figure 3). To determine the Cox threshold, we split the training set, performed principal components analysis in one half of the samples and used the model to predict survival in the other half. By varying the threshold of Cox scores and using 2-fold cross-validation, we found that a threshold of ±1.5 (averaged over five separate repeats of this procedure) best predicted survival (i.e., yielded the highest average partial log-likelihood ratio statistic).
Figure 3 Overview of the Strategy Used for the Development and Validation of a Prognostic Gene List
There were 340 transcripts (representing 259 genes whose Cox score equaled or exceeded this threshold), and they were used to perform principal components analysis on the entire training set. (For a full list of transcripts with unigene cluster ID, locus link ID, gene symbol, and gene ontology annotations, see Table S1.) As can be seen in Figure 4, only the first principal component was strongly correlated with survival. In 247 genes, high expression levels were associated with prolonged survival (Figure 4A), and in only 12 genes was high expression associated with shorter survival, including BAG2, DCBLD2, EDG2, GNAS, IGLC2, NCF1, NME2, PFN2, PRPS2, REG4, SLC7A5, and TFAP2C. There did not appear to be enrichment of gene ontology annotations in this prognostic gene set. However, some expression features suggested biological processes that underlie differences in tumor behavior. For instance, three genes involved in adhesion and diapedesis of lymphocytes, CD34, PECAM1, and VCAM1, show higher expression in tumors with good prognosis, and a lymphocytic-mediated immune response can alter the clinical course of cRCCs [4]. Furthermore, high expression of VCAM1 has been shown previously to predict survival in cRCC patients with metastatic disease [9]. Elevated expression of FZD2 and TCF4, members of the Wnt signaling pathway, also correlated with longer survival.
Figure 4 Outcome Prediction Using the SPC Risk Score
(A) Overview of the gene expression patterns of the 259 prognostic genes in the training set, with their SPC risk scores arranged in ascending order and the survival time in descending order. Each row represents a single gene, and each column a patient sample. The degree of color saturation corresponds to the ratio of gene expression in each sample compared to the mean expression across all samples.
(B) Gene expression profiles of the 259 prognostic genes in the test set.
(C) Kaplan-Meier estimates of disease-specific survival in low-, intermediate-, and high-risk groups of patients in the training set defined by the tertiles of SPC risk scores.
(D) Kaplan-Meier estimates of disease-specific survival of low, intermediate and high-risk groups of patients in the test set defined based on the tertiles of the SPC risk scores of the training set.
(E) Kaplan-Meier estimates of disease-specific survival in stage group I and II patients in the test set.
(F) Kaplan-Meier estimates of disease-specific survival in stage III and IV patients in the test set.
For each case, SPC analysis computed a risk score (SPC risk score) that represents the sum of the weighted expression levels for each of the 340 prognostic transcripts. Not surprisingly, the SPC risk score was highly correlated with survival in the training set (p < 0.001 by the log-rank test). To validate the SPC predictor, we computed risk scores for each of the 89 cases in the test set, using the model developed in the training set, and tested whether these scores were correlated with survival (Figure 4B). When the SPC risk score was used as a continuous variable, it was a strong predictor of survival in the independent test set (p < 0.001 by the log-rank test). Fewer genes from the SPC set also could predict outcome since genes were identified based on their correlation with survival. For instance, the top four genes in the SPC predictor could be used to predict survival in the test set at p = 0.02. Moreover, multivariate analysis showed that the SPC risk score provided powerful prognostication independent of stage, grade, and performance status (p < 0.001, by the Cox model likelihood ratio test) (Table 3). Further investigation showed that the log-relative risk was fairly linear in the SPC score. When cases in the test set were split into localized (stages I and II) and advanced disease (stages III and IV), SPC risk score as a continuous variable continued to be highly correlated with survival and was independent of grade and performance status (p = 0.011, and p = 0.045, respectively, by the Cox model likelihood ratio test).
Table 3 Prognostic Significance of SPC Risk Score Compared to Clinical Features (p-Values) by the Log-Rank Test Using the Test Set Samples
To illustrate the performance of the SPC risk score in predicting survival, we divided the training and test sets into tertiles based on the SPC risk scores of the training set. In both the training and test sets, Kaplan-Meier analysis showed that the group with the highest SPC risk score had significantly worse survival compared to the other two groups (Figure 4C and 4D). When used as a categorical predictor, the SPC risk score again predicted survival independent of grade, stage, and performance status in all tumors in the test set and in the high-stage tumors (stage groups III and IV; see Figure 4F), although not in the low-stage tumors (stage groups I and II; see Figure 4E), likely due to low numbers of high-risk cases in the test set. It should be emphasized, however, that the SPC risk score is continuous, and survival is directly correlated with the SPC risk score. Although cases can be assigned to risk categories based on the SPC risk score, outcome is better predicted when the SPC risk score is used continuously, rather than categorically (Figure 4C–4F).
Relationship of SPC Risk Score to the Gene Expression Subgroups
Most of the 259 prognostic genes comprising the SPC risk score were found in clusters that distinguished between the two major groups of cRCC that were defined by unsupervised hierarchical clustering analysis (i.e., between subgroups 1 and 2 and subgroups 3, 4, and 5) (see Figures 1A [black bar] and 2C). Despite this overlap, the SPC risk score predicted outcome independent of tumor subgroup in the test set (p = 0.0013, by the Cox model likelihood ratio test), even though the tumor subgroup had been assigned in the original hierarchical cluster of all 177 tumors (comprising both the training and test sets) and was based on differences in expression over 3,674 genes. Tumor subgroup, on the other hand, did not predict survival independent of the SPC risk score (p = 0.12 by the Cox model likelihood ratio test).
Discussion
We have identified gene expression patterns that correlate with survival after nephrectomy in cRCC. We used unsupervised hierarchical clustering analysis to identify five distinct subgroups that differed in their expression patterns over 3,674 genes. These subgroups were correlated with survival time after surgery that was independent of tumor stage, grade, and patient performance status. The consistency of gene expression within the subgroups, regardless of tumor stage, suggests that distinct molecular genetic changes present at early stages of tumor development determine the fate of the cancer and can be used to predict clinical outcome. The identification of these five new subgroups supports the use of gene expression profiling for prognostication in cRCC and highlights the value of unsupervised analytic methods to provide insights into the clinical and biological heterogeneity of cRCC.
We used a novel, semisupervised analytic strategy to identify 259 genes that better predicted survival than the gene expression subgroups, and we have validated this prognostic gene set on an independent group of patients. We used SPC analysis to compute a continuous risk score that predicted survival in the test set independent of stage, grade, performance status, and gene expression subgroup. Combining the SPC risk score with tumor grade, stage, and patient performance status may help identify patients with cRCC who have a high probability of being cured of their disease and need less intensive follow-up testing after surgery, or high-risk individuals who might be referred for adjuvant treatments even though their disease is clinically occult. The SPC method employed in this paper could be used generally in microarray studies to correlate gene expression with survival time. Other approaches, such as the model-based mixture proposal of Jones et al., could also be tried [23].
An interesting feature of the SPC prognostic gene set is that 95% of the genes show relatively high levels of expression in patients with good outcome and low expression in those with poor outcome. Notably, this pattern of gene expression was observed in a set of 51 prognostic genes identified in 29 cRCCs by Takahashi et al., and 15 of these genes were found in the SPC gene set [8]. Unfortunately, we were not able to evaluate the usefulness of the SPC gene set in predicting survival in their patients because their dataset is not publicly available. Another study by Vasselli and coworkers identified 45 genes associated with survival based on the Cox proportional-hazards score, using 58 stage IV tumors from patients with good performance status [9]. Those genes share minimal overlap (one out of 45) with the SPC gene set, possibly because this study included a highly selected group of patients with tumors of conventional and nonconventional histology. We and others have reported striking differences in transcript profiles of renal cancers of different histologies, and these differences could significantly influence the gene identified that correlate with prognosis [17,24].
The 259 prognostic genes do not show enrichment for any single biological pathway and are not localized to a single region of the genome. The diversity of the pathways represented in the SPC gene set argues that expression of different functional groups of genes contributes to cRCC growth, metastasis, and lethality.
Several molecules have been shown to correlate with prognosis in cRCC; however, most of them were not selected by SPC analysis as strong predictors of survival in our dataset [25,26]. While some of these molecules might have important biological roles in cRCC progression, they could be excluded from the SPC gene list because they are relatively weak predictors of survival. For instance, some transcripts, like ADFP (a hypoxia-induced gene) did not provide a strong enough correlation with survival to make the SPC gene set. However, in our dataset, ADFP was correlated with a number of the SPC genes in a cluster that defined main tumor groups I and II (see Figure 1) by unsupervised clustering analysis. When the 177 cases were separated into two groups based on their median expression level of ADFP, they displayed significantly different cancer-specific survival (p = 0.03). Carbonic anhydrase 9, another potential prognostic marker for advanced RCC showed uniformly low expression in subgroup 4, which has the worst survival rate, although its expression did not correlate with survival in the whole dataset. Therefore, predictions of outcome that are based on single genes will be less robust that that for multiple genes, such as those of the SPC predictor.
Gene expression profiling can improve outcome prediction in patients with cRCC beyond that provided by stage, grade, and patient performance status. Application of the SPC risk score in the clinical setting will depend on independent confirmation of our findings and could occur through custom DNA microarrays or quantitative reverse-transcriptase polymerase chain reaction assays [27–29]. Since as few as four genes in the SPC gene set can estimate prognosis, it should be possible to develop clinically useful predictors of survival based on these technologies. Regardless, our study demonstrates the molecular heterogeneity of cRCC and opens opportunities for improved biological understanding of the molecular subgroups of the disease and their response to therapy.
Supporting Information
Figure S1 Dendrogram Representing the Similarities in the Gene Expression Patterns between Experimental Samples
(451 KB PDF).
Click here for additional data file.
Table S1 The 259 Genes Predicting Survival Identified using SPC
(65 KB XLS).
Click here for additional data file.
Accession Numbers
The complete microarray dataset has been deposited in National Center for Biotechnology's Gene Expression Omnibus (http://www.ncbi.nlm.nih.gov/geo/) and is accessible through accession number GSE3538.
Patient Summary
Background
The kidneys filter the blood and eliminate waste in the urine through a complex system of filtration tubules. All of the blood in the body passes through the kidneys approximately 20 times an hour. Conventional renal cell carcinoma (cRCC) is the most common type of kidney cancer and arises from the cells that line the filtration tubules of the kidney. Nearly half of the people who get RCC die from the disease. Gene expression profiling, a laboratory technique, offers promise for guiding the diagnosis and treatment of cancers.
Why Was This Study Done?
The current method for estimating survival using clinical markers (such as tumor stage and grade) has limitations. As has been found for other cancer types, the hope is that gene expression profiling could identify molecular markers that could be used for more accurate diagnosis, prognosis, and possibly serve as drug targets for effective therapies. Several small gene expression studies of cRCC done so far have each identified prognostic gene sets, but these genes did not overlap, and studies have not been validated independently. This larger study looked systematically for variations in gene expression that were correlated with the clinical heterogeneity of cRCCs.
What Did the Researchers Do and Find?
The researchers studied a set of 177 tumors from patients for whom they had detailed clinical information, including data on long-term survival. They found a set of 259 genes whose activity in the tumor correlated with long-term survival independent of the standard clinical predictors. Most of the genes showed high levels of expression in patients with good outcome and low expression in those with poor outcome. They then used this information to show that they could accurately predict survival in an independent group of patients.
What Do These Findings Mean?
The researchers identified a set of genes whose activity predicted survival after surgery independent of clinical prognostic factors. This suggests that expression profiles could help to distinguish between more aggressive and less aggressive types of cRCCs. If confirmed by other studies, such expression profiles could be used with information on tumor grade, stage, and patient performance to help identify patients with cRCC who have a high probability of being cured and need less intensive treatment and follow-up testing after surgery and others whose cancers should be treated more aggressively.
Where Can I Get More Information Online?
The following Web sites have information on kidney cancer.
Cancer Research UK:
http://www.cancerresearchuk.org/
MedlinePlus:
http://www.nlm.nih.gov/medlineplus/ency/article/000516.htm
CancerBACUP:
http://www.cancerbacup.org.uk/Cancertype/Kidney
US National Cancer Institute:
http://www.cancer.gov/cancertopics/types/kidney
CancerLinks:
http://www.cancerlinks.com/kidney.html
Kidney Cancer Association:
http://www.kidneycancerassociation.org/
The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The authors would like to thank the patients who participated in this study.
Citation: Zhao H, Ljungberg B, Grankvist K, Rasmuson T, Tibshirani R, et al. (2006) Gene expression profiling predicts survival in conventional renal cell carcinoma. PLoS Med 3(1): e13.
Abbreviations
cRCCconventional renal cell carcinoma
RCCrenal cell carcinoma
SPCsupervised principal components
==== Refs
References
Vogelzang NJ Stadler WM Kidney cancer Lancet 1998 352 1691 1696 9853456
Jemal A Murray T Ward E Samuels A Tiwari RC Cancer statistics, 2005 CA Cancer J Clin 2005 55 10 30 15661684
Flanigan RC Salmon SE Blumenstein BA Bearman SI Roy V Nephrectomy followed by interferon alfa-2b compared with interferon alfa-2b alone for metastatic renal-cell cancer N Engl J Med 2001 345 1655 1659 11759643
Negrier S Escudier B Lasset C Douillard JY Savary J Recombinant human interleukin-2, recombinant human interferon alfa-2a, or both in metastatic renal-cell carcinoma N Engl J Med 1998 338 1272 1278 9562581
Frank I Blute ML Cheville JC Lohse CM Weaver AL A multifactorial postoperative surveillance model for patients with surgically treated clear cell renal cell carcinoma J Urol 2003 170 2225 2232 14634384
Patard JJ Kim HL Lam JS Dorey FJ Pantuck AJ Use of the University of California Los Angeles integrated staging system to predict survival in renal cell carcinoma: An international multicenter study J Clin Oncol 2004 22 3316 3322 15310775
Sorbellini M Kattan MW Snyder ME Reuter V Motzer R A postoperative prognostic nomogram predicting recurrence for patients with conventional clear cell renal cell carcinoma J Urol 2005 173 48 51 15592023
Takahashi M Rhodes DR Furge KA Kanayama H Kagawa S Gene expression profiling of clear cell renal cell carcinoma: Gene identification and prognostic classification Proc Natl Acad Sci U S A 2001 98 9754 9759 11493696
Vasselli JR Shih JH Iyengar SR Maranchie J Riss J Predicting survival in patients with metastatic kidney cancer by gene-expression profiling in the primary tumor Proc Natl Acad Sci U S A 2003 100 6958 6963 12777628
Boer JM Huber WK Sultmann H Wilmer F von Heydebreck A Identification and classification of differentially expressed genes in renal cell carcinoma by expression profiling on a global human 31,500-element cDNA array Genome Res 2001 11 1861 1870 11691851
Bullinger L Dohner K Bair E Frohling S Schlenk RF Use of gene-expression profiling to identify prognostic subclasses in adult acute myeloid leukemia N Engl J Med 2004 350 1605 1616 15084693
Gollub J Ball CA Binkley G Demeter J Finkelstein DB The Stanford Microarray Database: Data access and quality assessment tools Nucleic Acids Res 2003 31 94 96 12519956
Eisen MB Spellman PT Brown PO Botstein D Cluster analysis and display of genome-wide expression patterns Proc Natl Acad Sci U S A 1998 95 14863 14868 9843981
Bair E Hastie T Debashis P Tibshirani R Prediction by supervised principal components J Am Stat Assoc 2005 In press
Gnarra JR Tory K Weng Y Schmidt L Wei MH Mutations of the VHL tumour suppressor gene in renal carcinoma Nat Genet 1994 7 85 90 7915601
Ivan M Kondo K Yang H Kim W Valiando J HIFalpha targeted for VHL-mediated destruction by proline hydroxylation: Implications for O2 sensing Science 2001 292 464 468 11292862
Higgins JP Shinghal R Gill H Reese JH Terris M Gene expression patterns in renal cell carcinoma assessed by complementary DNA microarray Am J Pathol 2003 162 925 932 12598325
Perou CM Sorlie T Eisen MB van de Rijn M Jeffrey SS Molecular portraits of human breast tumours Nature 2000 406 747 752 10963602
Rosenwald A Wright G Chan WC Connors JM Campo E The use of molecular profiling to predict survival after chemotherapy for diffuse large B cell lymphoma N Engl J Med 2002 346 1937 1947 12075054
Ramaswamy S Ross KN Lander ES Golub TR A molecular signature of metastasis in primary solid tumors Nat Genet 2003 33 49 54 12469122
Bair E Tibshirani R Semi-supervised methods to predict patient survival from gene expression data PLoS Biol 2004 2 e108 10.1371/journal.pbio.0020108 15094809
Bair E Semi-supervised methods for predicting patient survival from microarray data [dissertation] 2004 Stanford (California) Stanford University
Ben-Tovim Jones L Ng SK Ambroise C Monico K Khan N Shoemaker JS Lin SM Use of microarray data via model-based classification in the study and prediction of survival from lung cancer Methods of microarray data analysis IV 2005 New York Springer 163 173
Schuetz AN Yin-Goen Q Amin MB Moreno CS Cohen C Molecular classification of renal tumors by gene expression profiling J Mol Diagn 2005 7 206 218 15858144
Yao M Tabuchi H Nagashima Y Baba M Nakaigawa N Gene expression analysis of renal carcinoma: Adipose differentiation-related protein as a potential diagnostic and prognostic biomarker for clear-cell renal carcinoma J Pathol 2005 205 377 387 15682440
Bui MH Seligson D Han KR Pantuck AJ Dorey FJ Carbonic anhydrase IX is an independent predictor of survival in advanced renal clear cell carcinoma: Implications for prognosis and therapy Clin Cancer Res 2003 9 802 811 12576453
van de Vijver MJ He YD van't Veer LJ Dai H Hart AA A gene-expression signature as a predictor of survival in breast cancer N Engl J Med 2002 347 1999 2009 12490681
Lossos IS Czerwinski DK Alizadeh AA Wechser MA Tibshirani R Prediction of survival in diffuse large-B-cell lymphoma based on the expression of six genes N Engl J Med 2004 350 1828 1837 15115829
Paik S Shak S Tang G Kim C Baker J A multigene assay to predict recurrence of tamoxifen-treated, node-negative breast cancer N Engl J Med 2004 351 2817 2826 15591335
|
16318415
|
PMC1298943
|
CC BY
|
2021-01-05 11:13:40
|
no
|
PLoS Med. 2006 Jan 6; 3(1):e13
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030013
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1631841110.1371/journal.pmed.0030019Research ArticleBioethicsScience PolicyEpidemiology/Public HealthHealth PolicyMedical EthicsEthicsHealth PolicyResearch MethodsSystematic reviews and meta-analysesAre Racial and Ethnic Minorities Less Willing to Participate in Health Research? Minorities in Health ResearchWendler David
1
*Kington Raynard
2
Madans Jennifer
3
Wye Gretchen Van
4
Christ-Schmidt Heidi
5
Pratt Laura A
3
Brawley Otis W
6
Gross Cary P
7
Emanuel Ezekiel
1
1Department of Clinical Bioethics, National Institutes of Health Clinical Center, National Institutes of Health, Bethesda, Maryland, United States of America2Office of Behavioral and Social Sciences Research, National Institutes of Health, Bethesda, Maryland, United States of America3National Center for Health Statistics, Centers for Disease Control and Prevention, Hyattsville, Maryland, United States of America4 Department of Epidemiology, Yale University School of Medicine, New Haven, Connecticut, United States of America5Statistics Collaborative, Washington, D. C., United States of America6Winship Cancer Institute, Emory University, Atlanta, Georgia, United States of America7Section of General Internal Medicine, Yale University School of Medicine, New Haven, Connecticut, United States of AmericaGill Paramjit Academic EditorUniversity of BirminghamUnited Kingdom*To whom correspondence should be addressed. E-mail: [email protected]
Competing Interests: The authors have declared that no competing interests exist.
Author Contributions: DW, RK, and EE designed the study. DW, JM, GV, HC, LAP, CPG, and EE analyzed the data. DW, RK, JM, GV, HC, LAP, OWB, CPG, and EE contributed to writing the paper.
2 2006 6 12 2005 3 2 e1908 6 2005 18 10 2005 Copyright: © 2006 Wendler et al.2006This is an open-access article distributed under the terms of the Creative Commons Public Domain declaration, which stipulates that, once placed in the public domain, this work may be freely reproduced, distributed, transmitted, modified, built upon, or otherwise used by anyone for any lawful purpose.
Minority Participation in Health Research-Facts and Fiction
Why are Minority Ethnic People Under-Represented in US Research Studies?
Background
It is widely claimed that racial and ethnic minorities, especially in the US, are less willing than non-minority individuals to participate in health research. Yet, there is a paucity of empirical data to substantiate this claim.
Methods and Findings
We performed a comprehensive literature search to identify all published health research studies that report consent rates by race or ethnicity. We found 20 health research studies that reported consent rates by race or ethnicity. These 20 studies reported the enrollment decisions of over 70,000 individuals for a broad range of research, from interviews to drug treatment to surgical trials. Eighteen of the twenty studies were single-site studies conducted exclusively in the US or multi-site studies where the majority of sites (i.e., at least 2/3) were in the US. Of the remaining two studies, the Concorde study was conducted at 74 sites in the United Kingdom, Ireland, and France, while the Delta study was conducted at 152 sites in Europe and 23 sites in Australia and New Zealand. For the three interview or non-intervention studies, African-Americans had a nonsignificantly lower overall consent rate than non-Hispanic whites (82.2% versus 83.5%; odds ratio [OR] = 0.92; 95% confidence interval [CI] 0.84–1.02). For these same three studies, Hispanics had a nonsignificantly higher overall consent rate than non-Hispanic whites (86.1% versus 83.5%; OR = 1.37; 95% CI 0.94–1.98). For the ten clinical intervention studies, African-Americans' overall consent rate was nonsignificantly higher than that of non-Hispanic whites (45.3% versus 41.8%; OR = 1.06; 95% CI 0.78–1.45). For these same ten studies, Hispanics had a statistically significant higher overall consent rate than non-Hispanic whites (55.9% versus 41.8%; OR = 1.33; 95% CI 1.08–1.65). For the seven surgery trials, which report all minority groups together, minorities as a group had a nonsignificantly higher overall consent rate than non-Hispanic whites (65.8% versus 47.8%; OR = 1.26; 95% CI 0.89–1.77). Given the preponderance of US sites, the vast majority of these individuals from minority groups were African-Americans or Hispanics from the US.
Conclusions
We found very small differences in the willingness of minorities, most of whom were African-Americans and Hispanics in the US, to participate in health research compared to non-Hispanic whites. These findings, based on the research enrollment decisions of over 70,000 individuals, the vast majority from the US, suggest that racial and ethnic minorities in the US are as willing as non-Hispanic whites to participate in health research. Hence, efforts to increase minority participation in health research should focus on ensuring access to health research for all groups, rather than changing minority attitudes.
Minorities in the US are as willing as non-Hispanic whites to participate in health research. Efforts to increase minority participation should focus on ensuring equal access to health research rather than changing minority attitudes.
==== Body
Introduction
To ensure the generalizability of research results, it is important that all groups participate in health research [1–4]. However, many commentators claim that racial and ethnic minority groups, especially in the US, are less willing to participate in health research [5–13]. While the US population includes an increasing percentage of individuals from minority groups, non-Hispanic whites still compose a majority of the population (Figure 1). It is widely believed that racial and ethnic minority groups in the US, especially African-Americans, are less willing than non-Hispanic whites to participate in health research. Many commentators believe that this relative unwillingness traces to past abuses, especially the notorious Tuskegee Syphilis Study [14–18], described as “the singular reason behind African-American distrust of the institutions of medicine and public health” [19].
Figure 1 Ethnic and Racial Composition of the United States
Data from year 2003.
The claim that racial and ethnic minority groups in the US are less willing to participate in health research seems to be validated by data showing that minority groups are underrepresented in at least some health research studies [20–25]. Yet, willingness to participate is just one factor that influences whether individual patients and patient groups participate in health research [9,26]. Other factors include whether they are informed of research opportunities, whether they are medically eligible to participate, and whether their personal circumstances, including child care demands, job flexibility, and geographic proximity to research sites, allow them to participate.
Simply assuming that minority groups' underrepresentation in some health studies is a result of their being less willing to participate may focus efforts aimed at increasing their participation on changing minority attitudes. If, however, minorities are equally willing to participate, and their lower participation in some studies traces to other factors, these efforts may prove ineffective, or even counterproductive. The assumption that minority groups are less willing to participate in health research also may inadvertently increase stigmatization, suggesting that minority groups are unwilling to bear their fair share of the burdens required to improve medical care.
A few studies have assessed the willingness of racial and ethnic minority groups to participate in individual research trials and trials that focus on single diseases [7,27,28]. However, we could find no published empirical data on the actual consent rates of minority groups for health research in general. To evaluate the widespread claim that racial and ethnic minorities are less willing to participate in health research, we assessed whether individuals from minority groups who were eligible and invited to participate in health research consented to enroll less frequently than non-Hispanic whites.
Methods
Non-Intervention Studies
The Tuskegee Syphilis Study, thought to be a major reason, particularly in the US, behind racial and ethnic minority groups' presumed unwillingness to participate in health research, was a US government-funded, epidemiologic study. To assess racial and ethnic minority groups' willingness to participate in current US government-funded, epidemiologic studies, we evaluated the health surveys conducted by the US National Center for Health Statistics (NCHS) for the most recent year for which data were available, year 2000.
The NCHS, the nation's principal health statistics agency, conducts two ongoing population based health surveys—the National Health Interview Survey (NHIS) and the National Health and Nutrition Examination Survey (NHANES). The NCHS, in collaboration with the National Immunization Program, also conducts the National Immunization Survey (NIS), which collects data that can be used to determine consent rates by race. The other health surveys conducted by the NCHS did not qualify for analysis, either because they are based on the NHIS sample, or because they are based on administrative records, not direct contact with individuals. Thus, three survey or non-intervention studies conducted by the NCHS that provide data on consent rates by race or ethnicity are included in the analysis.
The NHIS is an annual, in-person, household interview, designed to produce data representative of the civilian, non-institutionalized population of the US [29]. The sample consists of approximately 106,000 persons, in approximately 43,000 households in over 300 primary sampling units. The first part of the survey collects basic demographic and health data on all members of the household. A sample adult and child are then selected from each household to complete a more detailed interview that assesses illness, health-care utilization, and socioeconomic and demographic factors. All households selected into the sample are visited by interviewers to introduce the study and conduct a brief interview to determine eligibility. Data on race and ethnicity are collected at this initial contact. The NHIS invites a higher percentage of African-Americans and Hispanics to participate than the percentage of these two groups in the US population. Interview response rates were calculated for sample adults for whom demographic information was obtained during the initial contact. The NHIS neither conducts public outreach nor provides financial incentives.
The NIS is an annual, random-digit-dialing telephone survey of approximately 34,000 US households with at least one child 19–35 mo old [30]. After answering questions about the resident child's immunization status, participants are asked for approval to contact providers to obtain the child's immnunization records. The NIS neither conducts public outreach nor provides financial incentives.
The NHANES is an annual nationally representative sample survey of approximately 5,000 non-institutionalized US civilians. Extensive media and public outreach are conducted in each community to familiarize potential participants with the survey. The household interview component of the NHANES assesses respondents' health, health-care utilization, and demographic characteristics [31]. Participants do not receive any financial incentives for the interview portion of the NHANES.
Individuals who complete the interview portion of the NHANES are invited to participate in an extensive medical examination, which requires a separate consent. The medical examination lasts approximately 4 h and includes a physical examination and a second interview. The physical examination collects blood and urine samples for laboratory tests such as cholesterol levels, blood lead levels, and levels of exposure to other environmental health hazards. The interview includes questions on physical health, mental health, sexual behavior, and drug use. Participants who complete the physical examination receive $70–$100 for their time and effort. Because individuals must have already consented to the NHANES interview to be invited to participate in the medical examination, the consent rates for the medical examination are listed (Table 1), but not included in the overall statistical analysis.
Table 1 Interviews and Non-Intervention Studies
Health Intervention Studies
Because there are no databases of health intervention trials, we conducted a comprehensive literature search of published trials. Using PubMed (http://www.ncbi.nlm.nih.gov/entrez/), we searched 30 unique combinations of terms and strings of terms related to enrollment, refusal, race, and ethnicity in phase I trials, phase I/II trials, phase II trials, and randomized controlled trials (see Table S1 for search terms). Studies were eligible for inclusion if they documented the race or ethnicity of eligible individuals invited to enroll, as well as the race or ethnicity of those who actually enrolled.
Two authors (G. V. and C. P. G.) reviewed the titles of all 1,681 articles identified by the search terms, then retrieved abstracts for the 1,106 articles that included any terms related to consent rates, including “consent,” “eligible,” “refusal,” or “enrollment” (see Figure S1 for a description of the selection process). The abstracts of all 1,106 retrieved articles were reviewed for any terms or phrases suggesting that they might include data by race or ethnicity, including “race,” “ethnicity,” “Caucasian,” “African-American,” “Hispanic,” and “non-Hispanic white.” The full texts of the 68 articles that included any of these terms were reviewed to determine whether they included data on consent rates by race or ethnicity, yielding 17 unique articles.
Next, Web of Science was used to search for authors whose names appeared in the citations of two or more of the 17 identified articles, yielding 371 articles (see Table S2). Using the same selection process, the bibliographies of the 17 identified articles were reviewed for any articles mentioning consent or participation rates, yielding another 89 articles. The same two authors (G. V. and C. P. G.) repeated the selection process to search the original 1,681 articles to determine whether any of these additional 467 articles included data on consent rates by race or ethnicity. This search yielded no additional articles.
Finally, to assess whether our search missed any studies that document consent rates by race or ethnicity, we evaluated the articles published for an entire year for two different types of trials. First, we reviewed all randomized controlled trials published during the 1-y period beginning April 1, 1999, in four major clinical journals: Annals of Internal Medicine, JAMA, The Lancet, and The New England Journal of Medicine. Because we wanted to assess individual patients' willingness to enroll in health research, trials were included only if they used individual patients as the unit of randomization (as opposed to hospital, region, etc.). This search identified 172 articles. We next conducted a MedLine search to identify all phase I oncology trials that used safety as an endpoint, published in English in the year 2002. This search identified 250 articles.
Using the same search process that was used for the original 1,681 studies, all 422 so identified articles were reviewed to determine whether any documented consent rates by race or ethnicity. This search yielded one study that documented consent rates by race, which had been identified previously by our original MedLine search. The fact that this search of the published articles for an entire year for two different types of intervention trials did not yield any new articles suggests that our original search likely identified all studies published in English that documented consent rates by race or ethnicity.
Statistical Analysis
We defined the consent rates for a given study as the number of individuals in each reported racial or ethnic group who agreed to participate in the study divided by the number of individuals in that group who were invited to participate. The identified studies classified minority groups in four different ways: (1) African-Americans and Hispanics classified separately; (2) only African-Americans classified; 3) African-Americans classified separately and all other minorities grouped together; 4) all minorities grouped together. Because the original studies used different race/ethnic classifications, the minority group(s) to which non-Hispanic whites are compared in the present analysis varies across the studies.
For each study, we calculated an odds ratio (OR) and the associated 95% confidence interval (CI) using non-Hispanic whites as the reference group. The OR specifies, for each study, whether the reported minority group was more or less likely to consent to enrollment than non-Hispanic whites. An OR greater than one indicates that the minority group was more likely to consent than non-Hispanic whites; an OR of less than one indicates that the minority group was less likely to consent. A DerSimonian–Laird random effects model was used to estimate the summary OR and 95% CI for each type of study (i.e., non-intervention studies, clinical intervention studies, and surgical intervention studies), again using non-Hispanic whites as the reference group [32].
We considered the summary OR to be statistically significant if the 95% CI did not include one. We also used the following statistic to test for statistical significance:
where
is the estimated summary OR and s
* is the estimated standard deviation of the log summary OR. Under the null hypothesis that Λ = 1, X has a χ2 distribution with one degree of freedom. We also tested for homogeneity by calculating a Breslow–Day χ2 statistic.
Results
We found 20 health research studies that included sufficient data to determine consent rates by race or ethnicity. Eighteen of the twenty studies were single-site studies conducted exclusively in the US or multi-site studies where the majority of sites (i.e., at least 2/3) were in the US. Of the remaining two studies, the Concorde study was conducted at 74 sites in the United Kingdom, Ireland, and France, while the Delta study was conducted at 152 sites in Europe and 23 sites in Australia and New Zealand. Taken together, these 20 studies reported the enrollment decisions of over 70,000 individuals, the vast majority of whom were from the US, for a broad range of health research studies, from interviews and non-intervention studies to drug treatment and surgical trials.
For the three interview or non-intervention studies, African-Americans had a nonsignificantly lower overall consent rate than non-Hispanic whites (82.2% versus 83.5%; OR = 0.92; 95% CI 0.84–1.02; Table 1; Figure 2). For these same three studies, Hispanics had a nonsignificantly higher overall consent rate than non-Hispanic whites (86.1% versus 83.5%; OR = 1.37; 95% CI 0.94–1.98; Figure 3). Additionally, there was a significant lack of homogeneity for both comparisons (i.e., the test of homogeneity of the OR was rejected). A lack of homogeneity indicates that the relative willingness to enroll of minority groups versus non-Hispanic whites was not consistent, but varied significantly within the group of studies.
Figure 2 Comparison of African-American versus non-Hispanic White Consent Rates
Circle diameter is proportional to the sample size of the individual studies. The diamond represents the overall OR. The vertical line indicates the 95% confidence interval on the OR. Blue indicates interview and non-intervention studies; red indicates clinical intervention studies.
Figure 3 Comparison of Hispanic versus non-Hispanic White Consent Rates
Circle diameter is proportional to the sample size of the individual studies. The diamond represents the overall OR. The vertical line indicates the 95% confidence interval on the OR. Blue indicates interview and non-intervention studies; red indicates clinical intervention studies.
For the ten clinical intervention studies, African-Americans had a nonsignificantly higher overall consent rate than non-Hispanic whites (45.3% versus 41.8%; OR = 1.06; 95% CI 0.78–1.45; Table 2; Figure 2). Again, there was a significant lack of homogeneity among the ORs, indicating that the relative willingness to enroll of minority groups versus non-Hispanic whites varied significantly within the group of studies. For these same ten studies, Hispanics had a statistically significant higher overall consent rate than non-Hispanic whites (55.9% versus 41.8%; OR = 1.33; 95% CI 1.08–1.65; Figure 3).
Table 2 Clinical Intervention Trials
Table 3 reports the consent rates for the seven surgical intervention studies, which categorized all minority groups together. For these seven trials, minorities as a group had a nonsignificantly higher overall consent rate than non-Hispanic whites (65.8% versus 47.8%; OR = 1.26; 95% CI 0.89–1.77; Figure 4). While the test of homogeneity was only nominally significant (p = 0.046), six of the seven ORs were greater than one.
Figure 4 Comparison of Minority versus non-Hispanic White Consent Rates in Surgical Intervention Trials
Circle diameter is proportional to the sample size of the individual studies. The diamond represents the overall OR. The vertical line indicates the 95% confidence interval on the OR.
Table 3 Surgical Intervention Trials
Importantly, of the 20 studies identified, seven offered enrollment to very few minority individuals. For example, the BARI study of percutaneous transluminal coronary angioplasty versus coronary artery bypass graft for coronary artery disease offered enrollment to 3,832 non-Hispanic whites, but to only 16 individuals from all minority groups combined (Table 3). Similarly, the CASS study of surgery versus medical management for angina pectoris offered enrollment to 2,065 non-Hispanic whites, but to only 30 individuals from all minority groups (Table 3).
Discussion
We identified 20 health research studies that reported the consent rates by race or ethnicity of over 70,000 individuals, the vast majority of whom were from the US. These 20 studies reveal small differences in the rates at which non-Hispanic whites and minorities agree to participate in health research. Indeed, where there are differences in consent rates, individuals from minority groups tend to be slightly more willing to participate in health research, particularly for clinical and surgical intervention studies. These findings contradict the widely held view that racial and ethnic minority groups in the US are less willing than non-Hispanic whites to participate in health research.
Our findings are striking given that they represent the enrollment decisions of over 70,000 individuals, including over 14,000 individuals who were invited to participate in clinical and surgical intervention trials. Furthermore, although we found only 20 studies that reported consent rates by race or ethnicity, these studies represent a broad range of invasiveness and risk, from in-person interviews and medical chart reviews, to drug treatment and surgical trials. These studies also cover a broad range of conditions, including recurrent throat infection, substance abuse, schizophrenia, HIV infection, cancer, and cardiac diseases.
Studies suggest that various factors, including historic abuses like the Tuskegee study, may have undermined minority groups' trust in medical research, as measured by survey questions and focus groups [33,34]. These factors may have increased individuals' suspicions or decreased their level of trust. However, the present analysis reveals that these factors have not resulted in racial and ethnic minorities in the US being less willing to participate in health research [35].
Although we found only small differences in consent rates by race or ethnicity, we did find substantial differences by race and ethnicity in the number of individuals invited to participate. In particular, seven of the 17 clinical and surgical intervention studies offered enrollment to relatively few individuals from minority groups, substantially fewer than one would expect based on the percentage of the population composed of minority groups and the incidence of the diseases being studied. For instance, the CASS study of surgery versus medical management for angina pectoris offered enrollment to a total of 2,095 individuals, 2,065 of whom were non-Hispanic whites and only 30 of whom were from all minority groups combined. Yet, as of 1980, 17% of the U.S. population belonged to a minority group, and the estimated prevalence of angina pectoris is higher in minority groups, especially African-Americans and Hispanics, than in non-Hispanic whites [36]. Recognizing that this rough estimate of minority representation in the US fails to take into account other relevant considerations, US demographics and the prevalence of angina pectoris suggest that the CASS study, which recruited individuals in the late 1970s, should have offered enrollment to approximately 356 individuals from minority groups (17% of 2,095), more than ten times the 30 individuals from minority groups actually offered enrollment [37].
Looking at the number of individuals who participated, one might conclude that these studies support the thesis that minorities are less willing to participate in health research in the US. This conclusion is contradicted by the studies' actual consent rates. In the BARI study, individuals from minority groups agreed to participate at a significantly higher rate than non-Hispanic whites (62.5% versus 47.6%). Similarly, individuals from minority groups agreed to participate at a significantly higher rate than non-Hispanic whites in the CASS study (43.3% versus 37.1%).
We found a significant lack of homogeneity for all pooled statistics, with the exception of Hispanics' and non-Hispanic whites' comparative willingness to enroll in clinical intervention trials. This lack of homogeneity indicates that the relative willingness to enroll of minority groups versus non-Hispanic whites varies significantly within the various groups of studies.
The lack of homogeneity suggests that comparative willingness to enroll in specific studies cannot be inferred simply from the type of study, or the racial or ethnic groups in question. Instead, it appears that individuals from minority groups are more willing to enroll in some studies, and non-Hispanic whites are more willing to enroll in others. This finding suggests that willingness to enroll often is more a function of the characteristics of individual studies than a function of racial or ethnic identity. Hence, in cases where a study has difficulty enrolling individuals from a particular group, whether a minority group or non-Hispanic whites, it will be important to assess whether particular characteristics of the study account for this difference. For example, choice of study site may have an important impact on which groups are likely to enroll. Also, the lack of homogeneity suggests that it may be important to conduct further research to determine which characteristics of studies influence the willingness of racial and ethnic groups to participate.
Numerous writers have emphasized the need to increase minority participation in health research [38–42]. Such efforts are important for reasons of justice, and to ensure research findings are generalizable to the entire population. These efforts are especially important given data that minority groups are not represented adequately in some clinical trials [1–4,25]. If efforts to increase minority participation are to succeed, it is vital to understand why minority groups are underrepresented in some research trials. Widespread discussion of past abuses, and racial and ethnic minorities' presumed unwillingness to participate, has focused attention on the attitudes of individuals from minority groups. However, the current data suggest individuals from minority groups, at least in the US, are as willing as non-Hispanic whites to participate in health research when eligible and invited to participate.
This finding suggests that any underrepresentation of minority groups in health research, when it occurs, is likely the result of other factors, such as the fact that some studies invite comparatively few individuals from minority groups to participate [43]. Consequently, efforts to increase minority participation in health research should focus on increasing minority access to research participation, not changing minority attitudes [44–47]. To be successful, these efforts should take into account a number of considerations [48]. Informing minority groups of specific trials and inviting them to participate is an obvious step. In addition, health research trials should try to include sites that are accessible to minority groups, and identify and attempt to address factors that may undermine minority groups' participation in particular, such as the need for child care and reimbursement for travel expenses. Language barriers also may pose difficulties with recruiting some minority groups [49].
Several limitations suggest the need for future research. First, the current findings are limited to published articles that documented consent rates by race or ethnicity. Second, the vast majority of the over 70,000 individuals in the present analysis were from the US. The willingness of minority groups from other countries to participate in health research may differ from the willingness of minority groups in the US. For example, time and cost constraints may preferentially reduce the willingness of individuals from minority groups to participate in health research in general. Yet, this factor may be outweighed in the US by the fact that health care is not guaranteed, and individuals from minority groups may be more likely to use participation in research as a way to obtain access to physicians and health care.
Third, we did not assess minority groups' attitudes toward health research. The current findings do not rule out the possibility that past abuses have resulted in individuals from minority groups being more distrustful of health research than non-Hispanic whites. It may be that past abuses have led to greater distrust among minority groups, but that other factors result in individuals from minority groups being equally willing to participate overall. For instance, some minority groups are more likely to be from lower socioeconomic groups, and individuals from lower socioeconomic groups may be comparatively more willing to participate in research for a number of possible reasons, including a stronger sense of social obligation or to gain access to health treatments.
Fourth, our comprehensive search focused on clinical intervention trials, specifically phase I, phase I/II, phase II, and randomized controlled trials. Our search did not include prevention trials and natural history studies. Individuals from minority groups may be less willing than non-Hispanic whites to participate in these types of studies.
It is widely believed that racial and ethnic minorities are less willing to participate in health research. Such claims often focus on the US, where it is believed that minority groups' relative unwillingness to participate in health research traces to historic abuses, especially the notorious Tuskegee Syphilis Study. We found that racial and ethnic minorities in the US, particularly African-Americans and Hispanics, are as willing to participate, and in some instances more willing to participate, in health research than non-Hispanic whites, when eligible and invited to participate. These findings suggest that efforts to remedy any underrepresentation of minority groups in health research should focus on ensuring equal access to health research for all groups, not on changing attitudes. Efforts to increase minority groups' access to clinical research studies should focus on a range of considerations, including inviting minority groups to participate, using sites accessible to minority groups, and identifying and attempting to address factors that may undermine the participation of individuals from minority groups, such as the need for child care or reimbursement of travel expenses.
Supporting Information
Figure S1 Selection Process of Clinical Intervention Studies Identified by PubMed Search
(26 KB DOC).
Click here for additional data file.
Table S1 PubMed Search Terms
(63 KB DOC).
Click here for additional data file.
Table S2 Web of Science Search Terms Citation Search
(44 KB DOC).
Click here for additional data file.
Patient Summary
Background
Health research is meant to determine the best strategies for preventing and treating disease and to inform health policy. Approval of new drugs and health guidelines is usually issued at a national level. Many countries have ethnically and racially diverse populations, and we know that health parameters are not the same for the different groups. To make sure that health policies serve a diverse population, it is important that all ethnic and racial groups participate in health research.
Why Was This Study Done?
Several studies have found that minority groups, especially in the US, are often underrepresented in research studies. One possible explanation that has been suggested is that because of past abuses (especially of African-Americans in the notorious Tuskegee Syphilis Study), minorities are less willing to participate in medical research. The authors of this study wanted to test whether this was indeed the case.
What Did the Researchers Do and Find?
They looked through the health literature in a systematic way to find all recent studies that reported consent rates by race or ethnicity (every participant in health research has to give “informed consent”). They found 20 such studies, 18 of which were conducted primarily or exclusively in the US, covering a broad range of research from interview-based surveys to clinical trials. Taken together, these studies reported the decision of over 70,000 individuals who were invited to participate. The researchers then compared the consent rates (i.e., the proportion who actually agreed to participate and gave consent) among non-Hispanic whites, African-Americans, and Hispanics. They found very small differences in the overall willingness of minorities to participate in health research compared with non-Hispanic whites. However, they did find that many of the studies invited fewer minority individuals than would be representative for the US patient population.
What Does This Mean?
These results suggest that racial and ethnic minority groups, at least in the US, are as willing as non-minority individuals to participate in health research, but that they are underrepresented among the invited participants. Efforts to increase minority participation should therefore focus on offering participation to more minority individuals rather than focusing on changing minority attitudes. It will be important to determine why minorities are underrepresented among people invited to participate in health research. Another interesting question not answered by this study is what motivates individuals from the different groups to accept an invitation and participate in health research both in general and in a particular survey or trial.
Where Can I Find More Information Online?
Information on the Tuskeege study:
http://www.cdc.gov/nchstp/od/tuskegee/
http://www.pbs.org/newshour/bb/health/may97/tuskegee_5-16.html
http://www.npr.org/programs/morning/features/2002/jul/tuskegee/commentary.html
Office of Minority Health Affairs of the US National Heart, Lung, and Blood Institute:
http://www.nhlbi.nih.gov/about/omha/
Pages on minority resources and initiatives at the US National Human Genome Research Institute:
http://www.genome.gov/10011199
Report on Inclusion of Women and Minorities in Research from the Office for Protection from Research Risks:
http://www.hhs.gov/ohrp/humansubjects/guidance/hsdc94-01.htm
Thanks to Joel Verter for assistance with the statistical analysis and three anonymous reviewers for their helpful comments on a previous version of the manuscript. The opinions expressed are the authors' own. They do not represent any position or policy of the National Institutes of Health, the Centers for Disease Control and Prevention, or the Department of Health and Human Services. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Citation: Wendler D, Kington R, Madans J, Van Wye G, Christ-Schmidt H, et al. (2006) Are racial and ethnic minorities less willing to participate in health research? PLoS Med 3(2): e19.
Abbreviations
CIconfidence interval
NCHSNational Center for Health Statistics
NHANESNational Health and Nutrition Examination Survey
NHISNational Health Interview Survey
NISNational Immunization Survey
ORodds ratio
==== Refs
References
Hussain-Gambles M Ethnic minority under-representation in clinical trials: Whose responsibility is it anyway? J Health Organ Manag 2003 17 138 143 12916177
Britton A McKee M Black N McPherson K Sanderson C Threats to applicability of randomized trials: Exclusion and selective participation J Health Serv Res Policy 1999 4 112 121 10387403
Allmark P Should research samples reflect the diversity of the population? J Med Ethics 2004 30 185 189 15082815
Freedman LS Simon R Foulkes MA Friedman L Geller NL Inclusion of women and minorities in clinical trials and the NIH Revitalization Act of 1993—The perspective of NIH clinical trialists Control Clin Trials 1995 16 277 285 8582146
Shavers VL Lynch CF Burmeister LF Factors that influence African-Americans' willingness to participate in medical research studies Cancer 2001 91 Suppl 233 236 11148585
Green BL Partridge EE Fouad MN Kohler C Crayton EF African-American attitudes regarding cancer clinical trials and research studies: Results from focus group methodology Ethn Dis 2000 10 76 86 10764133
Thompson EE Neighbors HW Munday C Jackson JS Recruitment and retention of African American patients for clinical research: An exploration of response rates in an urban psychiatric hospital J Consult Clin Psychol 1996 64 861 867 8916612
Dennis BP Neese JB Recruitment and retention of African American elders into community-based research: Lessons learned Arch Psychiatr Nurs 2000 14 3 11 10692801
Shavers-Hornaday VL Lynch CF Burmeister LF Torner JC Why are African Americans under-represented in medical research studies? Impediments to participation Ethn Health 1997 2 31 45 9395587
Gauthier MA Clarke WP Gaining and sustaining minority participation in longitudinal research projects Alzheimer Dis Assoc Disord 1999 13 S29 S33 10369515
Svensson C Representation of American blacks in clinical trials of new drugs JAMA 1989 261 263 265 2909024
Williams CL Tappen R Buscemi C Rivera R Lezcano J Obtaining family consent for participation in Alzheimer's research in a Cuban-American population: Strategies to overcome barriers Am J Alzheimers Dise Other Demen 2001 16 183 187
Shavers VL Lynch CF Burmeister LF Racial differences in factors that influence the willingness to participate in medical research studies Ann Epidemiol 2002 12 248 256 11988413
El-Sadr W Capps L The challenge of minority recruitment in clinical trials for AIDS JAMA 1992 267 955
Reverby SM More than fact and fiction: Cultural memory and the Tuskegee syphilis study Hastings Cent Rep 2001 31 22 28 12974115
Thomas CR Pinto HA Roach M Participation in clinical trials: Is it state-of-the-art for African Americans and other people of color? J Natl Med Assoc 1994 86 177 182 8189450
Bonner GJ Miles TP Participation of African Americans in clinical research Neuroepidemiology 1997 16 281 284 9430127
Kirp DL Bayer R Needles and race Atlantic Monthly 1993 July 38 42
Gamble VN Under the shadow of Tuskegee: African Americans and health care Am J Public Health 1997 87 1773 9366634
Heiat A Gross CP Krumholz HM Representation of the elderly, women and minorities in heart failure clinical trials Arch Intern Med 2002 162 1682 1688 12153370
Mason S Hussain-Gambles M Leese B Atkin K Brown J Representation of South Asian people in randomized clinical trials: Analysis of trials' data BMJ 2003 326 1244 1245 12791739
Svensson CK Representation of American blacks in clinical trials of new drugs JAMA 1989 13 263 265
Hall WD Representation of blacks, women and the very elderly (aged > or = 80) in 28 major randomized clinical trials Ethn Dis 1999 9 333 340 10600055
Killien M Bigby JA Champion V Fernandez-Repollet E Jackson RD Involving minority and underrepresented women in clinical trials: The National Centers of Excellence in Women's Health J Womens Health Gend Based Med 2000 9 1061 1070 11153102
Murthy VH Krumholz HM Gross CP Participation in cancer clinical trials: Race-, sex-, and age-based disparities JAMA 2004 291 2720 2726 15187053
Hussain-Gambles M Atkin K Leese B Why ethnic minority groups are under-represented in clinical trials: A review of the literature Health Soc Care Community 2004 12 382 388 15373816
Bartlett C Davey P Dieppe P Doyal L Ebrahim S Women, older persons, and ethnic minorities: Factors associated with their inclusion in randomized trials of statins 1990 to 2001 Heart 2003 89 327 328 12591845
Kressin NI Meterko M Wilson NJ Racial disparities in participation in biomedical research J Natl Med Assoc 2000 92 62 69 10800293
Centers for Disease Control and Prevention National Health Interview Survey (NHIS) 2005 Available: http://www.cdc.gov/nchs/nhis.htm . Accessed 26 October 2005
Centers for Disease Control and Prevention The National Immunization Survey 2005 Available: http://www.cdc.gov/nis/Default.htm . Accessed 26 October 2005
Centers for Disease Control and Prevention National Health and Nutrition Examination Survey 2005 Available: http://www.cdc.gov/nchs/nhanes.htm . Accessed 26 October 2005
DerSimonian R Laird N Meta-analysis in clinical trials Control Clin Trials 1986 7 177 188 3802833
Giselle CS Thomas SB St. George DM Distrust, race, and research Arch Intern Med 2002 162 2458 2463 12437405
Sugarman J Kass NE Goodman SN Perentesis P Fernandes P What patients say about medical research IRB 1998 20 1 7
Brown DR Topcu M Willingness to participate in clinical treatment research among older African Americans and whites Gerontologist 2003 43 62 72 12604747
American Stroke Association, American Heart Association Heart disease and stroke statistics—2003 update 2002 Dallas American Heart Association Available: http://www.americanheart.org/downloadable/heart/10590179711482003HDSStatsBookREV7-03.pdf . Accessed 26 October 2005
Hobbs F Stoops N US Census Bureau (2002 November) Demographic trends in the 20th century. Census 2000 special reports, series CENSR-4 Washington (D. C.) US Government Printing Office Available: http://www.census.gov/prod/2002pubs/censr-4.pdf . Accessed 26 October 2005
Gifford AL Cunningham WE Heslin KC Andersen RM Nakazono T Participation in research and access to experimental treatments by HIV-Infected patients N Engl J Med 2002 346 1373 1382 11986412
King TE Racial disparities in clinical trials N Engl J Med 2002 346 1400 1402 11986416
Sateren WB Trimble EL Abrams J Brawley O Breen N How sociodemographics, presence of oncology specialists, and hospital cancer programs affect accrual to cancer treatment trials J Clin Oncol 2002 20 2109 2117 11956272
Haynes MA Smedley BD The unequal burden of cancer: An assessment of NIH research and programs for ethnic minorities and the medically underserved 1999 Washington (D. C.) National Academy Press 352
Underwood SM Alexander GG Participation of minorities and women in clinical cancer research Ann Epidemiol 2000 10 S1
Simon MS Du W Flaherty L Philip PA Lorusso P Factors associated with breast cancer clinical trials participation and enrollment at a large academic medical center J Clin Oncol 2004 22 2046 2052 15082724
Lee JY Marks JE Simpson JR Recruitment of patients to cooperative group clinical trials Cancer Clin Trials 1980 3 381 384 7428142
Hunter CP Frelick RW Feldman AR Bavier AR Dunlap WH Selection factors in clinical trials: results from the Community Clinical Oncology Program Physicians' Patient Log Cancer Treat Rep 1997 71 559 565
Appel LJ Vollmer WM Obarzanek E Aicher KM Conlin PR Recruitment and baseline characteristics of participants in the Dietary Approaches to Stop Hypertension trial J Am Diet Assoc 1999 99 S69 S75 10450297
Brawley O Tejeda H Minority inclusion in clinical trials issues and potential strategies J Natl Cancer Inst Monogr 1995 17 55 57 8573455
Swanson GM Ward AJ Recruiting minorities into clinical trials: Toward a participant-friendly system J Natl Cancer Inst 1995 87 1747 1759 7473831
Hunt S Bhopal R Self reports in research with non-English speakers BMJ 2003 327 352 353 12919965
Robinson D Woerner MG Pollack S Lerner G Subject selection biases in clinical trials: Data from a multicenter schizophrenia treatment study J Clin Psychopharmacol 1996 16 170 176 8690832
McKay JR Alterman AI Snider EC O'Brian CP Effect of random versus nonrandom assignment in a comparison of inpatient and day hospital rehabilitation for male alcoholics J Consult Clin Psychol 1995 63 70 78 7896993
Gorkin L Schron EB Handshaw K Shea S Kinney MR Clinical trial enrollers vs. nonenrollers: The CAST and REACT project Control Clin Trials 1996 17 46 59 8721801
Rimer BK Schildkraut JM Lerman C Lin TH Audrain J Participation in a women's breast cancer risk counseling trial. Who participates? Who declines? High Risk Breast Cancer Consortium Cancer 1996 77 2348 2355 8635106
Corbie-Smith G Viscoli CM Kernan WN Brass LM Sarrel P Influence of race, clinical, and other socio-demographic features on trial participation J Clin Epidemiol 2003 56 304 309 12767406
Kaluzny A Brawley O Garson-Angert D Shaw J Godley P Assuring access to state-of-the-art care for U.S. minority populations: The first 2 years of the Minority-Based Community Clinical Oncology Program J Natl Cancer Inst 1993 1 1945 1950
Moore DA Goodall RL Ives NJ Hooker M Gazzard BG How generalizable are the results of large randomized controlled trials of antiretroviral therapy? HIV Med 2000 1 149 154 11737343
Westerberg VS Miller WR Tonigan JS Comparison of outcomes for clients in randomized versus open trials of treatment for alcohol use disorders J Stud Alcohol 2000 61 720 727 11022812
Diener-West M Earle JD Fine SL Hawkins BS Moy CS The COMS randomized trial of iodine 125 brachytherapy for choroidal melanoma, II: Characteristics of patient enrolled and not enrolled Arch Ophthalmol 2001 119 951 965 11448318
CASS Investigators Coronary artery surgery study (CASS): A randomized trial of coronary artery bypass surgery. Comparability of entry characteristics and survival in randomized patients and nonrandomized patients meeting randomization criteria J Am Coll Cardiol 1984 3 114 128 6361099
Paradise JL Bluestone CD Bachman RZ Colborn DK Bernard BS Efficacy of tonsillectomy for recurrent throat infection in severely affected children. Results of parallel randomized and nonrandomized clinical trials N Engl J Med 1984 310 674 683 6700642
Williford J Krol WF Buzby GP Comparison of eligible randomized patients with two groups of ineligible patients: Can the results of the VA Total Parenteral Nutrition clinical trail be generalized? J Clin Epidemiol 1993 46 1025 1034 8263575
Marcus SM Assessing non-consent bias with parallel randomized and nonrandomized clinical trials J Clin Epidemiol 1997 50 823 828 9253394
King SB Barnhart HX Kosinski AS Weintraub WS Lembo NJ Angioplasty or surgery for multivessel coronary artery disease: Comparison of eligible registry and randomized patients in the EAST trial and influence of treatment selection on outcomes Am J Cardiol 1997 79 1453 1459 9185632
Hochman JS Sleeper LA Webb JG Sanborn TA White HD Early revascularization in acute myocardial infarction complicated by cardiogenic shock N Engl J Med 1999 341 625 631 10460813
Feit F Brooks MM Sopko G Keller NM Rosen A Long-term clinical outcome in the Bypass Angioplasty Revascularization Investigation Registry: Comparison with the randomized trial Circulation 2000 101 2795 2802 10859284
|
16318411
|
PMC1298944
|
CC0
|
2021-01-05 11:14:05
|
no
|
PLoS Med. 2006 Feb 6; 3(2):e19
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030019
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0030028SynopsisInfectious DiseasesMicrobiologyEpidemiology/Public HealthGastroenterology/HepatologyHealth PolicyInfectious DiseasesEpidemiologyGastroenterologyTravel MedicineMedicine in Developing CountriesAdjusting Cholera Models to Recent Experimental Data Synopsis1 2006 6 12 2005 3 1 e28Copyright: © 2006 PLoS Medicine.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Hyperinfectivity: A Critical Element in the Ability of
V. cholerae to Cause Epidemics?
==== Body
Infectious disease modeling has a long history, going back to at least Daniel Bernouli's smallpox model from 1760. The discipline is driven by the desire to understand the dynamics of an outbreak or epidemic in order to plan control strategies. The hope is that better models will also allow prediction of future outbreaks, and inform not just preparedness but also prevention strategies. One critical component of all infectious disease models—and by some experts seen as the bottleneck of most models—is the mode of transmission. Interaction between modelers and experimentalists who study transmission from environment to humans and transmission between humans is therefore crucial.
In 2002, Andrew Camilli and colleagues reported that passage through a human host potentiated the infectivity of the cholera pathogen
Vibrio cholerae (Nature 417: 642–645). They isolated and characterized
V. cholerae from human stools, and found that passage through the human gastrointestinal tract induces a transient hyperinfectious state in the bacteria, which is perpetuated for a number of hours, even outside the human host. (Infectiousness was determined in competition assays in infant mice.) This hyperinfectious state was associated with specific gene expression profiles that differed from those of bacteria in their normal aquatic reservoirs.
Scanning electron micropgaph of
Vibrio cholerae
The study caught the attention of David Hartley and colleagues, who saw a chance to improve modeling in the field of cholera epidemics. Hartley was interested because Camilli's results shed new light on a fundamental question in cholera epidemiology, namely, what is the relative importance of human-to-human (i.e., fecal to oral) versus environment-to-human infection (through contaminated food or water)? If the infective dose of bacteria that have become hyperinfectious because of recent passage through a human host is lower than that of environmental
V. cholerae, this would support a crucial role of human-to-human transmission in the generation of cholera epidemics.
Hartley and colleagues found that incorporation of the existence of a hyperinfectious state into their disease models resulted in a much better fit of the predictions with the observed explosive epidemic patterns of past cholera outbreaks. On one hand, this result lends theoretical support for Camilli's results and suggests that his findings in laboratory animals have clinical relevance. On the other hand, it strongly suggests that human-to-human transmission is crucial for cholera epidemics and pandemics, and that health measures must focus on minimizing risk of transmission of the short-lived hyperinfectious form of
V. cholerae that results from transmission through a human host. Finally, there is the intriguing possibility that similar hyperinfectious states exist for other bacteria, something that seems well worth exploring.
|
0
|
PMC1298945
|
CC BY
|
2021-01-05 11:13:40
|
no
|
PLoS Med. 2006 Jan 6; 3(1):e28
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030028
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0030034SynopsisCancer BiologyGenetics/Genomics/Gene TherapyPharmacology/Drug DiscoveryOncologyDrugs and Adverse Drug ReactionsGeneticsNew Approach to Combat Glioblastoma Shows Promise in Preclinical Studies Synopsis1 2006 6 12 2005 3 1 e34Copyright: © 2006 PLoS Medicine.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
EGF Receptor-Targeted Synthetic Double-Stranded RNA Eliminates Glioblastoma, Breast Cancer, and Adenocarcinoma Tumors in Mice
==== Body
Glioblastoma multiforme (GBM) develops in the tissue of the brain and grows quickly, often becoming very large before a person experiences symptoms and is diagnosed. Surgery is done to remove as much of the tumor as possible, and followed with radiation and/or chemotherapy to slow progression of the disease. But despite aggressive treatment, the cancer almost inevitably recurs, and patients usually die within a year.
Researchers are studying several ways to treat GBM using gene therapy. Some approaches target healthy cells to enhance their ability to fight cancer; others target cancer cells to destroy them or prevent their growth. Researchers are also experimenting with immune therapies, which seek to restore and stimulate the body's natural ability to recognize and attack cancer cells.
In general, scientists agree that an effective treatment for GBM must have several components. These include being highly selective to avoid damage to non-cancerous brain tissue and efficient cell killing, preferably by simultaneous activation of multiple killing mechanisms. Moreover, as it is unlikely that any treatment can target all cancer cells, it will be necessary to achieve a cancer-specific “bystander effect,” which kills neighboring tumors but not normal cells.
Alexander Levitzki and colleagues tested a therapy that tries to meet all these criteria by taking advantage of the frequent (50%–70%) overexpression of the epidermal growth factor receptors (EGFRs) in GBM. In those tumors, the number of EGFRs on tumor cells is 10–20 times higher than that on non-tumor cells. Based on this difference, they reasoned that any treatment that targets the EGFR would reach predominantly tumor cells. They built a non-viral delivery vehicle that could center in on cells expressing the EGFR and trigger internalization of the complex of the receptor and the bound vehicle. As a “toxic cargo,” Levitzki and colleagues selected double-stranded RNA (polyinosine-cytosine, poly IC). Double-stranded RNA doesn't occur naturally in eukaryotic cells, but is often associated with viral infections. To protect themselves against “viral takeover,” multicellular organisms have evolved efficient defense mechanisms that result in apoptotic death of cells that contain dsRNA.
Glioblastoma in treated (right) and control (left) animals
The team found that the poly IC induced rapid apoptosis in the target cells in vitro and in glioblastomas in mice. The therapy induced complete regression of pre-established intracranial tumors in nude mice, with no obvious adverse toxic effects on normal brain tissue. And one year later, the treated mice were still disease-free and healthy.
EGFR overexpression is common in other cancer types as well, and further in vivo experiments in mice showed that non-viral delivery of poly IC completely eliminated pre-established breast cancer and adenocarcinoma xenografts derived from EGFR overexpressing cancer cell lines, suggesting that the strategy might be applicable to other cancer types. The researchers also suggest that the principle of ligand-guided delivery of dsRNA at a particular receptor that is overexpressed and undergoes endocytosis might be applicable in a range of cancers.
Experience tells that curing cancer in mice is easy and curing cancer in humans is hard. It is too early to tell whether this is one of the few approaches that will still look promising after the next round of tests. In light of the encouraging results of this study and the lack of effective treatments for GBM, however, Robert Weil (DOI:
10.1371/journal.pmed.0030031) suggests in an accompanying commentary that the use of dsRNA delivered by non-viral vectors deserves to be fast-tracked to the clinic.
|
0
|
PMC1298946
|
CC BY
|
2021-01-05 11:13:40
|
no
|
PLoS Med. 2006 Jan 6; 3(1):e34
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030034
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0030035SynopsisBioinformatics/Computational BiologyCancer BiologyGenetics/Genomics/Gene TherapyMedical InformaticsNephrologyOncologyPathologyOncologyGeneticsMedical InformaticsPathologyRenal MedicineExpression Profiling Predicts Survival in Kidney Cancer Synopsis1 2006 6 12 2005 3 1 e35Copyright: © 2006 PLoS Medicine.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Gene Expression Profiling Predicts Survival in Conventional Renal Cell Carcinoma
==== Body
Nearly 95,000 people worldwide die from kidney cancer every year, and renal cell carcinoma (RCC) accounts for most of the deaths. Surgery can cure 60%–70% of patients with localized disease and prolong survival in patients with metastatic disease, but survival rates after treatment have not improved appreciably over the past 30 years.
Current survival estimates based on clinical characteristics such as tumor size and grade are not very accurate, and the varied response to surgery and other treatments suggests an underlying diversity that is not captured by the clinical parameters. As has been the case for other cancer types, researchers hope that comprehensive molecular genetic analysis will reveal distinguishing features that could serve as prognostic indicators, improve outcome prediction, and inform treatment decisions. Several previously reported expression-profiling studies of relatively small sets of RCCs suggested that comprehensive molecular genetic analysis might be useful in this cancer type as well.
To further the understanding of the genetics and molecular biology underlying RCC, James Brooks and colleagues determined the gene expression patterns of a set of 177 tumors from patients with detailed clinical information available, including long-term follow-up. Based on the results, the researchers could divide the tumors into five distinct subgroups, which differed in the expression patterns of over 3,000 genes. These subgroups correlated with survival after nephrectomy. The correlation was independent of tumor stage and grade, suggesting that molecular and genetic changes early during tumorigenesis determine the characteristics of a particular cancer and can be used to predict clinical outcome.
Brooks and colleagues then used a computational tool to identify 259 genes for which expression status was highly predictive of clinical outcome. The genes in this prognostic set represent a range of molecular pathways, and map to different parts of the genome. They found that 95% of them are expressed at high levels in tumors of patients with a good prognosis and at low levels in more aggressive cancers.
In an independent group of patients, the researchers used these 259 genes to calculate a risk score, and showed that it predicted patient survival independent of clinical parameters. They suggest that combining this risk score with tumor grade, stage, and patient performance status might help to identify patients with RCC who have a high probability of being cured and need less intensive adjuvant treatment and follow-up testing after surgery, as well as those who should receive more aggressive treatment.
A group of renal cell carcinomas share similar expression patterns
The next step will be to test the value of the risk score in independent studies. To be able to determine expression profiles in routine clinical settings, it will also be necessary to further reduce the number of genes in the prognostic set. Brooks et al. find that as few as four genes from their prognostic set can estimate outcome. They acknowledge that “application of the [supervised principal components] risk score in the clinical setting will depend on independent confirmation of [their] findings,” but conclude “that it should be possible to develop clinically useful predictors of survival based on these technologies.”
|
0
|
PMC1298947
|
CC BY
|
2021-01-05 10:40:33
|
no
|
PLoS Med. 2006 Jan 6; 3(1):e35
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030035
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0030037SynopsisInfectious DiseasesEpidemiology/Public HealthHealth EconomicsHealth PolicyHIV/AIDSSexual HealthChronic Disease ManagementHealth EconomicsHIV Infection/AIDSMedicine in Developing CountriesResource allocation and rationingHAART: A Cost-Effective Option for South Africa Synopsis1 2006 6 12 2005 3 1 e37Copyright: © 2006 PLoS Medicine.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Cost-Effectiveness of Highly Active Antiretroviral Therapy in South Africa
==== Body
There were an estimated 370,000 AIDS deaths in South Africa in 2003 alone. It is, therefore, not surprising that the apparent reluctance of the South African government to support the provision of antiretroviral treatment to people with HIV/AIDS has been the subject of much controversy internationally. The situation is, however, changing, and South Africa is now seeing a scaling up of access to highly active antiretroviral therapy (HAART) and a gradual reduction in HAART prices.
HAART, nevertheless, remains an expensive option, and one that many low-income countries are unable to afford. South Africa is better placed than most sub-Saharan nations to increase access to HAART, but it is clearly essential to establish how much this will cost the country's public health sector and what will be the benefits. It is unfortunate, therefore, that cost-effectiveness studies on HAART have so far been limited to the developed world.
In the January issue of
PLoS Medicine, Motasim Badri and colleagues publish a study of the use of the cost-effectiveness of HAART conducted in South Africa. During the study period (January 1995 to December 2000), HAART was not available in the publicly funded South African health-care sector.
The research was funded by Secure the Future—a Bristol-Myers Squibb initiative to provide resources for capacity building and for the search for sustainable interventions to address HIV/AIDS in sub-Saharan Africa— and took place in HIV clinics affiliated to the University of Cape Town. The sponsors had no involvement in the study design, analysis, or decision to publish. The study was based on a prospective cohort study—the Cape Town AIDS Cohort (CTAC).
The researchers compared the cost of services for 292 patients who were given HAART with the costs for a matched comparison group (with the same number of patients) who were not given any antiretroviral drugs. Twenty-seven patients in each group had AIDS; the others were HIV-infected but did not have AIDS. The researchers calculated costs per patient year (PPY) and per life-year gained (LYG)—i.e., the total cost divided by the number of extra years the treated patients lived. Calculations were done separately for patients with AIDS and for those without AIDS.
Patients on HAART required fewer hospital admissions. Depending on how long the patient survived and the price of antiretrovirals used, HAART reduced treatment costs for those patients with AIDS. For this group, the cost savings ranged from US$219–US$2,116. For patients without AIDS, the yearly cost of treatment (ranging from US$597–US$1,772) was, in the opinion of the authors, and after taking into account the South African standard of cost of living, considered to be affordable. However, it is expected that South Africa will soon be able to manufacture antiretroviral drugs locally and more cheaply. This would increase the amount saved by introducing HAART.
The study had a number of limitations. Because HAART was not used in routine clinical practice, the researchers had to compare a group of patients enrolled in clinical trials with a control group that was not part of the trials. The study was also confined to the use and cost of services; but when a person is infected with HIV and becomes ill or dies from AIDS, it is clearly not only the health services that face costs. The patient, their family, and the country suffer financially. HAART, as a more effective treatment might also lower these “indirect” costs, but this was not an issue examined here. It is to be hoped that further research includes an evaluation of the indirect costs and benefi ts. Nevertheless, the present study should encourage policymakers in low- and middle-income countries to consider introducing HAART into public-sector health care; reductions in the use of hospital services by patients with HIV could free scarce resources, to the benefi t of all who use the health services.
|
0
|
PMC1298948
|
CC BY
|
2021-01-05 11:13:41
|
no
|
PLoS Med. 2006 Jan 6; 3(1):e37
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030037
|
oa_comm
|
==== Front
PLoS MedPLoS MedpbioplosbiolPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0030040SynopsisBioethicsScience PolicyEpidemiology/Public HealthHealth PolicyMedical EthicsHealth PolicySystematic reviews and meta-analysesResearch MethodsEthicsMinority Participation in Health Research—Facts and Fiction Synopsis2 2006 6 12 2005 3 2 e40Copyright: © 2006 PLoS Medicine.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Are Racial and Ethnic Minorities Less Willing to Participate in Health Research?
==== Body
It is widely believed that racial and ethnic minority groups, especially in the US, are less willing to participate in health research than non-minority groups. According to this view, minority groups' comparative unwillingness to participate is due to a lack of trust in health research and health researchers, which traces to past abuses, particularly the Tuskegee Syphilis Study. Conducted from 1932–1972, the US government–funded Tuskegee study examined the natural course of syphilis. The participants, 399 African-American men in the late stages of syphilis, were enrolled and offered free medical care but kept in the dark about the nature of their illness and the purpose of the study. Participants were told that they were being treated for “bad blood,” but the doctors had no intention of curing them and even withheld penicillin treatment when it became available. When the experiment was finally ended—after public outcry following exposure in the media—28 of the men had died directly of syphilis, 100 were suffering from related complications, 40 of their wives had become infected, and 19 of their children had been born with congenital syphilis.
Given that the study was not halted until 1972, it would not be surprising if its memory influenced the attitudes of minority individuals toward health research today. This is a potential problem, because it is essential that participants in health research are as diverse as the population whose health should be improved as a result of the research. (And in the US today, one in five people is from a minority group). Only representative participation ensures that the outcomes of health research can be generalized to a diverse population.
Is participation in health research representative of the population? A number of studies suggest that in the US it is not; minority groups are often under-represented in US research studies. But what are the reasons? Are minority groups less willing to participate, or are they given fewer opportunities to participate? Answering this question is vital for efforts to increase minority participation in health research. Should these efforts focus on changing minority attitudes or on removing potential barriers to participation, such as whether minorities are adequately informed of research opportunities?
In a systematic review in this month's
PLoS Medicine, David Wendler and colleagues assessed whether individuals from minority groups who were eligible and invited to participate in health research were indeed less likely to consent to participate than non-minority individuals. The authors identified 20 health research studies that reported consent rates by race or ethnicity. Eighteen were single-site studies conducted exclusively in the US or multi-site studies conducted primarily or exclusively in the US. The 20 studies collectively report the enrollment decisions of over 70,000 individuals for a broad range of health research studies.
For the three non-intervention studies, African-Americans had a non-significantly lower overall consent rate than non-Hispanic whites, while Hispanics had a non-significantly higher overall consent rate than non-Hispanic whites. For the ten intervention studies, African Americans' overall consent rate was non-significantly higher than that of non-Hispanic whites, while Hispanics had a statistically significant higher overall consent rate than non-Hispanic whites. For the seven surgery trials, minorities as a group had a non-significantly higher overall consent rate than non-Hispanic whites.
Although Wendler and colleagues found only small differences in consent rates by race or ethnicity, they found substantial differences by race or ethnicity in the number of individuals invited to participate. For example, one study of medical versus surgical management of angina offered enrollment to 2,065 non-Hispanic whites but to only 30 individuals from minority groups.
This study, Wendler and colleagues say, “suggests that racial and ethnic minorities are as willing as non-Hispanic whites to participate in health research.” Indeed for some kinds of studies, minority individuals seem more willing to enroll than non-minority individuals. The authors acknowledge their study's limitations, particularly the fact that most individuals were from the US and the willingness of minority groups outside the US to participate in health research may be very different. Other important questions not addressed by this study are what motivates people to participate in health research, and whether motives differ between majority and minority groups. Despite these and other remaining questions, the results of this study suggest that efforts to increase minority participation in health research should concentrate on ensuring better access to health research for all groups, rather than on changing the attitudes of minority groups.
|
0
|
PMC1298949
|
CC BY
|
2021-01-05 11:14:05
|
no
|
PLoS Med. 2006 Feb 6; 3(2):e40
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030040
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 10.1371/journal.pmed.0030045SynopsisAllergy/ImmunologyRheumatologyRheumatologyImmunology and AllergyAutoimmune DiseasesCMV and Triggers of Systemic Sclerosis 1 2006 6 12 2005 3 1 e45Copyright: © 2006 PLoS Medicine.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
Antibodies against Human Cytomegalovirus in the Pathogenesis of Systemic Sclerosis: A Gene Array Approach
==== Body
Systemic sclerosis (SSc) is an autoimmune disease characterized by structural and functional vascular abnormalities, immunologic changes, and excessive extracellular matrix deposition, leading to fibrosis of the skin and the internal organs. The activation of the immune system is vital in the pathogenesis of SSc. Autoantibodies directed against cell surface antigens are believed to induce endothelial cell damage and apoptosis, which are thought to be a primary event in the pathogenesis of the disease, leading to a cascade of stimulatory changes involving many cells, including fibroblasts, T lymphocytes, and macrophages. In turn, these activated cells secrete substances, including cytokines and their soluble receptors, and enzymes and their inhibitors. These substances lead to changes in the extracellular matrix compounds, including fibronectin, proteoglycans, and collagen types I, III, V, and VII; specifically TGF-beta, interleukin-4 (IL-4), and platelet-derived growth factor are profibrotic cytokines.
Environmental factors may be involved in the disease's pathogenesis. Previous research has shown that a molecular mimicry mechanism links antibodies against the human cytomegalovirus (hCMV)-derived protein UL94 to the pathogenesis of SSc. The UL94 epitope shows homology with NAG-2, a surface molecule highly expressed on endothelial cells. Anti-UL94 peptide antibodies have been shown to induce apoptosis of endothelial cells upon engagement with the NAG-2-integrin complex.
Cytomegalovirus in the pathogenesis of systemic sclerosis
In a new study, Claudio Lunardi and colleagues examined further whether hCMV could be involved in the pathogenesis of fibrosis, and also attempted to dissect out the molecular mechanisms behind the disease. They found that NAG-2 was also expressed on dermal fibroblasts, and that anti-UL94 antibodies bind to fibroblasts.
Using gene arrays, they analyzed the transcriptional profile in endothelial cells and dermal fibroblasts in response to treatment with antibodies against the UL94 peptide. Exposure of endothelial cells to anti-UL94 antibodies had a profound impact on gene expression, resulting in the upregulation of 1,645 transcripts. The genes altered encoded for adhesion molecules, chemokines, colony-stimulating factors, growth factors, and molecules involved in apoptosis. Dermal fibroblasts showed an upregulation of 989 transcripts and acquired a “scleroderma-like” phenotype, with upregulation of genes involved in extracellular matrix deposition, growth factors, chemokines, and cytokines.
To confirm these findings, the investigators measured the levels of chemokines, cytokines, growth factors, and collagen type I in the supernatants of stimulated and unstimulated cells. They found that the concentration of the molecules was higher in the cells incubated with anti-hCMV antibodies, confirming that genes' upregulation was paralleled by the induction of protein synthesis.
Taking this analysis further into patients, the investigators measured the serum concentrations of some cytokines, chemokines, and adhesion molecules in patients and controls, and confirmed that the genes found overexpressed in vitro following stimulation with anti-hCMV antibodies could indeed be of relevance in vivo.
Altogether, these findings suggest that these cross-reacting antivirus antibodies were able to induce not only endothelial cell activation and apoptosis but also fibroblast activation. They could thus be a single trigger of the two hallmarks of SSc, vascular damage and fibrosis.
|
0
|
PMC1298950
|
CC BY
|
2021-01-05 10:40:33
|
no
|
PLoS Med. 2006 Jan 6; 3(1):e45
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030045
|
oa_comm
|
==== Front
BMC BioinformaticsBMC Bioinformatics1471-2105BioMed Central London 1471-2105-6-2681628007810.1186/1471-2105-6-268DatabaseMiMiR: a comprehensive solution for storage, annotation and exchange of microarray data Navarange Mahendra [email protected] Laurence [email protected] Derek [email protected] Vihar [email protected] Helen [email protected] Nicola [email protected] Fatimah [email protected] Justin [email protected] Peter [email protected] Helen C [email protected] CSC-IC Microarray Centre, Imperial College, Hammersmith Campus, DuCane Road, London W12 ONN, UK2005 9 11 2005 6 268 268 10 5 2005 9 11 2005 Copyright © 2005 Navarange et al; licensee BioMed Central Ltd.2005Navarange et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The generation of large amounts of microarray data presents challenges for data collection, annotation, exchange and analysis. Although there are now widely accepted formats, minimum standards for data content and ontologies for microarray data, only a few groups are using them together to build and populate large-scale databases. Structured environments for data management are crucial for making full use of these data.
Description
The MiMiR database provides a comprehensive infrastructure for microarray data annotation, storage and exchange and is based on the MAGE format. MiMiR is MIAME-supportive, customised for use with data generated on the Affymetrix platform and includes a tool for data annotation using ontologies. Detailed information on the experiment, methods, reagents and signal intensity data can be captured in a systematic format. Reports screens permit the user to query the database, to view annotation on individual experiments and provide summary statistics. MiMiR has tools for automatic upload of the data from the microarray scanner and export to databases using MAGE-ML.
Conclusion
MiMiR facilitates microarray data management, annotation and exchange, in line with international guidelines. The database is valuable for underpinning research activities and promotes a systematic approach to data handling. Copies of MiMiR are freely available to academic groups under licence.
==== Body
Background
Large amounts of microarray gene expression data are being generated by researchers attracted by the insights that can be gained from knowing the expression levels of thousands of genes [1]. In parallel, there has been a growing appreciation of the potential value of these data beyond the description found in most papers and for the need for well-annotated gene expression databases and mechanisms for data exchange [2]. Recognition of the role of standards in the management and use of microarray data has prompted a number of initiatives within the academic community [3,4]. The Microarray Gene Expression Data Society (MGED)[5], in particular, has made considerable progress towards providing guidelines and standards for the description, management and exchange of microarray data. MIAME, the Minimum Information About a Microarray Experiment, specifies the data content required to ensure that the data can be easily interpreted and the results independently verified [6]. MAGE provides a format for representation of all types of microarray gene expression data and includes both an object model (MAGE-OM) and a markup language (MAGE-ML) for data exchange [7]. The MGED Ontology supplies concepts, a controlled vocabulary and structure for describing the complete process of carrying out a microarray experiment [8-10]. Other groups within MGED are developing standards for data transformation, for other functional genomics technologies (the Reporting Structure for Biological Investigations Working Group) and are defining the Minimum Information Specification For In Situ Hybridization and Immunohistochemistry Experiments.
Support for these standards has come from multiple sources: manufacturers of hardware and software, the (US) Food and Drug Administration, which has endorsed MIAME as a standard for Voluntary Genomic Data submissions [11], and journals that have made the transfer of MIAME-compliant data into the public domain a precondition of publication [12-14]. There are now three public repositories for microarray data worldwide: ArrayExpress at the European Bioinformatics Institute [15-17], the Gene Expression Omnibus (GEO) at the National Cancer Institute [18,19] and the Center for Information Biology Gene Expression Database (CIBEX) at the DNA Data Bank of Japan [20,21].
Despite considerable advances in formats and standards for data content and description, and increased access to software and microarray data, few groups in academia are appropriately placed to take advantage of large-scale microarray data for high-throughput analysis. Some of the challenges lie in the generation of large data sets. Small perturbations in the microenvironment can have a large effect on the gene expression profile, so experimental manipulations need to be vigorously controlled and recorded [22]. The large number of interrelated variables make the data hard to annotate and only a few databases support the fine detail needed to make sense of these data [23,24]. As a result, many microarray facilities and academic groups lack resources for data management that support the existing standards and facilitate global data comparison and rigorous statistical analysis [2,25].
Here we describe MiMiR, an Oracle relational database, which was designed to assist users in tackling many of these challenges. This is a heavy-duty application used to collect, annotate, store and exchange microarray data and the related experimental information from thousands of arrays in a MAGE-ML compliant format. MiMiR has tools for data annotation that promote systematic use of the MGED and other ontologies, for importing data from the scanner and for exporting data to other MAGE-based databases. MiMiR has been optimised for fast and efficient data retrieval and high performance multi-user access.
Construction and content
Overview
MiMiR is a stand-alone application that was built to support work at an academic microarray centre. Staff at the Centre collaborate with researchers (users) to design experiments and to generate, manage, analyse, annotate and publish microarray data. MiMiR was therefore designed (i) to support flexible data collection and systematic annotation at a high level of detail, (ii) for data exchange with other MAGE-based databases and (iii) as the basis for a high-performance application for heavy-duty data analysis, mining and fast data retrieval by large numbers of concurrent users. These objectives have been achieved and the database is in constant use, although new functionality is continually being added via modular plug-ins.
MiMiR architecture
MiMiR is modular and consists of a fully integrated Oracle relational database, a custom-built Java user interface for data entry, management and the generation of reports, configuration software and a MAGE-ML export tool for data exchange with other MAGE-compliant databases (Figure 1a). The network database and web reporting application are server software, while the upload and export applications are client software.
Figure 1 Two views of MiMiR. Figure 1a shows information flow in and out of the database. Information about the experimental design, spots on the array (the sequence of each reporter), signal intensities for each coordinate on the scanned array and the derived spot level data are uploaded into the database. The same interface is also used for queries and to update the data. Information about the data in MiMiR is available via web-based reports. Data on individual experiments may be exported as MAGE-ML to databases such as ArrayExpress. Figure 1b shows the architecture of the database and the underlying applications. The BC4J framework drives interactions between the user interface and the data model. JClient mediates navigation of the database. XDK permits the creation of XML directly from MiMiR. Tomcat offers potential for remote access of the database by users on different sites.
Data stored in MiMiR include experimental annotation, the signal intensities for each coordinate on a scanned array, derived signal intensity data and the sequence present at each spot on the array. Data can be viewed using web-based reports, queried using the query interface and automatically exported as MAGE-ML.
MiMiR is a flexible application that was designed to support workflow at the Centre. The Centre operates on a service basis, generating data on large numbers of samples using a number of well-established quality control and standard operating procedures and primarily uses Affymetrix GeneChips. The staff annotate data on the biomaterial and treatments for researchers, frequently after hybridisation and scanning (information on the array design is obtained from the manufacturers and signal intensity data is loaded as part of the application). In consequence, the table structure and user interface were built to support flexible data capture, permitting data from multiple hybridisations to be loaded at once and linked to an experiment at a later stage. The structure was optimised for Affymetrix data and is built around standard operating protocols and quality controls, many of which are now used in other centres that generate microarray data using the GeneChip platform [26]. Information can easily be viewed, updated or modified using the same screens as those for data entry and there is provision for remote access.
The MAGE object model provides a comprehensive specification for representation of all types of microarray data and has more than 100 tables, many with only one or two columns [7]. Although data storage is well served by such a structure, the object model was not designed for efficient data retrieval, and many tables have to be traversed in order to access information. While remaining true to the data content requirements of MAGE-OM, MiMiR was built using a denormalised, purely relational, design in which columns and tables containing related information were combined. The design includes some deviations from the MAGE-OM: MiMiR permits multiple measurement units for each treatment in the Biomaterial Package, whereas MAGE-OM permits only one. Deviations such as this offer flexibility and save annotators time without compromising function. The table structure was designed so that many queries involve only the information in a single table. Querying is also facilitated by indexing the data using a few repeated primary keys that also link the tables. User-defined data types not supported by many simpler database platforms were avoided to enhance cross (database) platform interoperability. The use of a restricted number of array types with common design features meant that some of the tables in MAGE were not needed. The resulting table structure has only 35 tables.
Oracle-specific features were used to (a) improve data retrieval from large tables, (b) enhance data integrity and (c) obtain a modular application that is easy to maintain and navigate. Data in large tables were partitioned onto multiple disks and the partitions indexed such that data from the largest table containing raw signal intensities (37 GB) can be retrieved in less than 15 seconds (we use 4 disks). Data integrity is enhanced with constraints, sequences and indexes, e.g., triggers automatically check that references to data are in place upon data entry and that information uploaded, or modified, is updated in all the relevant records. SQL scripts regularly check for and report anomalies. Applications such as the BC4J framework provide tight integration between the user interface and the data model (Figure 1b).
The user interface
Overview
A single interface is used for data entry, modifying records and querying the data. The screens are intuitive to use and are based around nine modules that represent the packages in MAGE. The modules provide a format for capturing information on all aspects of the experiment, from the researcher to the experimental design, arrays used, the biological sample, reagents (including catalogue numbers, suppliers and batch numbers), protocols, references, treatments, preparation of the labelled extract, hybridisation and scanning. Free text boxes provide space for recording additional information.
The interface was built using Oracle J Developer, which provides a sophisticated development environment tightly integrated with the database and application server [27]. The application is currently deployed as a .jar file and can be run on any PC that has JRE1.4.0.1, so that large numbers of individual users can access MiMiR from their desktops. The Business Components for Java (BC4J) framework (Oracle) is a tool for creating scalable high-performance J2EE applications and is implemented within JDeveloper. The software allows developers to build a logical persistence layer on which applications can be built and tested using Java objects. BC4J has two major advantages. Firstly, it mediates the interaction between the user interface and the data model (Figure 1b), freeing the developer from having to write, test and debug infrastructure code because applications only deal with the persistent data via Java objects. Secondly, it permits the creation of reusable objects and features 'Patterns' that encapsulate 'best practices' in J2EE coding e.g., Model-Viewer-Controller. The use of BC4J means that changes in the interface and database behaviour are easy to implement and modify and the database requires little maintenance.
JClient, used in the user interface, permits comprehensive navigation of MiMiR using predefined 'classes', or controls, while JSP and Tomcat provide potential for remote data retrieval. Data may be exported using XSQL and the GDAC Exporter [28]. Alternatively, XDK may be used to write XML directly from the database without the need for additional files. This makes XML-based data retrieval and export fast and efficient and requires minimal code.
The user interface mimics the logical flow of information to be recorded from the organism to the data generated from the scan (Figure 2). Data is typically generated before annotation, so MiMiR was built to be 'hybridisation-centric', as opposed to 'experiment-centric'. Annotation of experimental data can be initiated at any module and linked to the relevant scan at a later stage. Objects are tracked using unique identifiers, some of which are also used to link modules (e.g., the Array and Hybridisation modules are linked via the ChipID). Gene expression (scan) data are uploaded using an automatic function and labelled extracts are linked with the gene expression data in the Hybridisation module. Help is available in the form of pop-up windows for each field in the data entry application. There is also a comprehensive user manual [29] and a web-based annotation tutorial [30]. Quality control (QC) flags can be appended at a number of stages (total RNA, cRNA and scan) and are typically used for identifying problems in sample processing and to assist in data analysis.
Figure 2 Data captured in MiMiR. Diagram showing the information captured from the details of the experiment through to the information gained from the scanned array and how they are linked. Although arrows indicate the logical flow of information capture as the experiment proceeds, data entry can be initiated at any of the boxes containing bold text and connected to other information about the experiment at a later stage.
Data annotation and MiMiR Ontology Viewer (MOV)
Annotation of experiments is important because the transcriptome is highly dynamic and changes rapidly in response to small perturbations [31]. However, detailed information about microarray experiments frequently remains in the laboratory notebooks of researchers and is not recorded in a form that is readily usable. Capturing this information is difficult because the variables that affect gene expression, and thus data interpretation, are not always known or accurately recorded. The use of ontologies can enhance the quality, accuracy and consistency of annotation, increasing the utility of the data for analysis and mining. However, there were no comprehensive ontologies, or controlled vocabularies, available until recently that could be used to describe this information systematically, and identifying the most appropriate terms and learning how to use ontologies can be challenging. The MiMiR data entry interface was therefore built with tools to encourage consistent annotation and the use of ontologies, to reduce problems associated with misspelling and different units and to minimise the use of free text. Where necessary, annotators can create their own terms and definitions. The MGED Ontology Viewer, or MOV, facilitates use of the MGED Ontology [8-10]. MOV displays the ontology as a dynamic class tree that can be searched and navigated (inset, Figure 3) and includes definitions for individuals. MOV is used to 'drill down' to locate terms in the ontology. Once the term has been identified and selected, the class and instance are automatically placed in the relevant fields in MiMiR, along with the 'MO:' designation that identifies the source of the term. Drop-down menus with predefined lists are used wherever possible. MOV can readily be used with any ontologies available in the DAML format and is based on SAX (Simple API for XML). Other ontologies, such as the NCI Metathesaurus [32] are also frequently used at the Microarray Centre. Examples of data annotation using MiMiR are available at [30].
Figure 3 A snapshot of the user interface showing how experimental factors are captured. A snapshot of the data entry screen showing how the 'factors' (the variables) in an experiment are captured within MiMiR. The experiment involved treatment of cells in culture with 0, 0.04, 0.1 or 0.2 mM hydrogen peroxide and harvesting at 0, 2, 4 or 8 hours. The lower part of the screen shows the MiMiR Ontology Viewer (MOV) with the MGED Ontology displayed as a tree in the pane on the left. Definitions of terms highlighted in the tree are displayed on the right. The Factor Category and Value columns at the top of the screen were populated using MGED Ontology terms selected from MOV. The measurement kind, measurement unit type and measurement unit were selected from drop-down menus and the Factor Name and the Measurement Value were entered manually.
The repeat function permits the data associated with an individual sample to be replicated. The application tracks the data across multiple tables and creates new records in each of the relevant tables respecting all primary and foreign key constraints. Individual fields can be modified to reflect minor differences among records. These functions ensure that a high level of detail can be captured, including samples that are pooled or split, with minimal typing and saves time when annotating large datasets.
Automated data upload
Expression data are loaded using an automated upload function and linked to the experimental information using a unique hybridisation identifier. Spot-level data typically load slowly, so data upload was enhanced using asynchronous processing. Files sent to MiMiR are automatically detected by the server, which creates six loader files, one for each of the tables storing spot-level data. The system administrators are emailed to inform them that upload occurred successfully or of the error logs to be checked.
Formats for viewing the data
A web reporting utility and reports screens provide summary information based on predefined queries, such as the amount of data in MiMiR and the types of arrays run. These are primarily used for checking data integrity.
Query tool
The query tool supports user-defined searches composed of single or multiple search criteria acting on single or multiple fields. The query result pages have 'clickable' screens that allow the user to navigate through records pulled out by a particular search.
Export of data as MAGE-ML
Export of data from MiMiR to other MAGE-based databases is carried out by converting the relevant data into MAGE-ML. Most of the data in MiMiR is sent to ArrayExpress upon publication and so an ArrayExpress export tool was developed. The export process involves retrieval of information about the experiment and the related biosources, biosamples, etc. using Oracle XSQL scripts. XML is generated directly from SQL statements run against the database. Affymetrix output files (.CEL, .CHP, and .EXP) are converted to MAGE-ML using the Affymetrix GDACExporter software [28]. The outputs from the XSQL and GDAC Exporter are merged using the eXtensible Stylesheet Language (XSL) to produce a single MAGE-ML file, validated and then transferred to the ArrayExpress ftp site.
Utility and discussion
MiMiR provides a comprehensive data management environment and was built to support work at a busy academic microarray centre. A problem-based approach to design led to the creation of a purely relational database based on MAGE. MiMiR was built to capture and track information about experiments, as it becomes available, for exchange of data with other databases and provides the foundation for an application that supports higher level data analysis (Figure 1a). The database was designed as a high performance application and features a denormalised relational table structure, indexed primary keys that link the tables, data partitioned on multiple disks and an asynchronous data upload tool. Asynchronous data upload sets priorities and loads data when resources are available, which saves time and offers flexibility: an Affymetrix .CEL file of 15 MB typically loads in 10 seconds.
Most of the work carried out at the Microarray Centre is done on a service basis and the data entry interface was therefore built to support data upload via multiple access points (Figure 2). Information on each experiment is entered directly, at the level of detail chosen by the data annotator, and usually exceeds that required by MIAME. Details of researchers collaborating with the Centre and the title and description of the experiment are usually entered first; in other cases, the experimental data generated by hybridisation and scanning are entered first and the data annotated afterwards. Data can be updated using the same screens as those for data entry.
Figure 3 shows one of the 'Experiment' screens for a project, carried out in collaboration with the Centre [33]. Hydrogen peroxide at 0, 0.04, 0.1 or 0.2mM was added to rat neonatal myocytes in culture and cells were harvested at 0, 2, 4 or 8 hours after treatment. The 'factors', or intended experimental variables, are the 'compound' hydrogen peroxide and the 'timepoints' at which the samples were harvested. MGED Ontology terms have the 'MO:' prefix. Terms of the predefined 'Measurement Kind', 'Measurement Unit Type' and 'Measurement Unit', were chosen from dropdown menus. MOV was used to select the Factor Category and Values from the BiologicalFactorCategory class from a hierarchical view of the ontology. The terms highlighted by the annotator are automatically pasted into the relevant fields in the data entry screen. Here, text was only entered manually in the Factor Name and Measurement Value columns.
Although the treatment of the cells with hydrogen peroxide is unique to this experiment, many of the subsequent stages, in which total RNA was prepared and the samples were labelled, fragmented, hybridised to arrays and scanned, differ little among experiments and are carried out using standard operating protocols [26]. In consequence, data entries for the latter stages of sample preparation were linked to the commonly used protocols. The data were annotated using fields for recording sample and experiment-specific details such as batch numbers of reagents, deviations in protocols, settings, yields and quality control information. The experimental information for one sample was completed, the whole record was duplicated using the repeat function and then modified for subsequent samples.
The data was submitted to ArrayExpress using the export tool and can be accessed from [34] (E-MIMR-12). Although the database is currently only used for MAGE export, MAGE import can be carried out using Oracle's XQuery tool [35]. A parser is only needed for situations where the MAGE imported is systematically different from that generated in house, for example, where information stored in a field in MiMiR is coded as free text.
Comparison with other systems
Many groups in the microarray data generating community recognise the importance of large scale databases for data exchange, annotation and management and there are now three public repositories of microarray data all of which accept gene expression data from multiple platforms. GEO uses the Simple Omnibus Format in Text format, while ArrayExpress and CIBEX use MAGE for data exchange. Some of the larger databases that support academic communities are the Stanford Microarray Database [36-38] that holds array data from the spotted array platform and the RNA Abundance Database [24,39,40] that holds data from multiple platforms.
MiMiR shares similarities to RAD in that both support discrete academic communities generating microarray data, are based on MAGE, and have been built for efficient querying and data retrieval, both provide extensive tools for data annotation and the use of ontologies and support export of data to ArrayExpress. BASE and MARS are two other comprehensive microarray management systems [41,42]. Like MiMiR they are also designed to support large numbers of users and to capture information usually found in a LIMS. Unlike MiMiR, they permit the user to carry out data transformation and to store the intermediate results of analysis, however, they have less extensive tools for data annotation and the systematic use of ontologies. BASE accepts Affymetrix, two colour and comparative genome hybridisation arrays, while MARS is built for multicolour arrays only. MiMiR also shares features with The Public Expression Profiling Resource (PEPR) [43]. Both of these Oracle databases are built around the Affymetrix platform and support data generation at microarray centres.
Future work
Planned additions to MiMiR include extension of the object model to accommodate data from multiple applications of spotted arrays and the annotation associated with clinical samples. Development of a web-based portal integrating multiple data analysis tools and integration of the database with Rosetta Resolver [44] are currently underway.
Conclusion
MiMiR is an Oracle relational database for microarray data storage, annotation, export and analysis that supports MAGE and MIAME compliance. The application is used for data management across a large and dispersed user-base and draws heavily on MGED standards and data structures.
Advantages of using MiMiR include:
• Data structure optimised for fast and accurate data retrieval, supporting large numbers of users (our implementation contains data from more than 2000 hybridisations (more than 70 GB)).
• Flexible user interface that permits data entry as data accumulates.
• Extensive focus on accurate and consistent sample annotation and data tracking. Including features for recording detailed experimental information, viewing and using ontologies, repeating records, generating web-based reports and querying the data.
Despite differences in the microarray gene expression databases and platforms available, support for common standards, the ability to share data and the willingness of investigators to put their data into the public domain are enhancing the utility of all data. This, in turn, is yielding greater insights into the biological systems underlying these data and predicts a strong future for knowledge gained from use of array-based technologies.
Availability and requirements
Project Name: MiMiR
Project Homepage:
Server Operating System: MS Windows (2000/XP/Server2003). MiMiR has a modular design and has been developed using standard tools such it can be modified to run under other operating systems and database environments. However, we do not anticipate that the entire system can be ported as a unit because features such as the Perl upload services, asynchronous data processing, GDAC Exporter and BC4J are platform-specific.
Client (User Interface) Operating System: All 32-bit MS Windows (95/98/NT/2000/XP)
Database Requirements: Oracle 9i
Programming Language: Java, Perl.
Copies of MiMiR are freely available to academic groups under license. Further information and requirements using the database management system can be obtained from [29].
Authors' contributions
MN and HC designed and conceived of the database. MN carried out the high level design, developed and implemented MiMiR. LG worked on comprehensive data annotation, management and coordination of the later stages of the project. MN, LG and HC drafted the manuscript. DF developed the user interface, the web reports and worked to enhance the Ontology Viewer. LG, HB, NC, FR and PB generated and annotated microarray data and populated and tested the database. VW developed the first versions of the Ontology Viewer. JH developed the ArrayExpress export tool. HC guided and coordinated execution of the project. All authors participated in optimising the design of MiMiR and have read and approved the final manuscript.
Acknowledgements
The authors thank the MGED community, Angela Clerk, Tim Kemp, Tim Aitman, Chris Higgins, Helen Parkinson, Anna Farne, Tim Rayner, Helen Figueira, Stephanie Dillman, Kumaran Mandek, Philippe Rocca-Serra, the Medical Research Council for funding and the reviewers for helpful comments and testing the system.
==== Refs
The Chipping Forecast II Supplement to Nature Genetics 2002 32
Bassett DE Eisen MB Boguski MS Gene expression informatics - it's all in your mine Nature Genetics 1999 21 51 55 9915502
Brazma A Editorial: On the importance of standardisation in life sciences Bioinformatics 2001 17 113 114 11238066
Stoeckert CJJ Causton HC Ball CA Microarray Databases: standards and ontologies Nature Genetics Supplement: The Chipping Forecast II 2002 32 469 473
Microarray Gene Expression Data Society
Brazma A Hingamp P Quackenbush P Sherlock G Spellman P Stoeckert C Aach J Ansorge W Ball CA Causton HC Glenisson P Holstege FCP Kim I Markowitz V Matese JC Robinson A Sarkans U Stewart J Taylor R Vilo J Vingron M Minimum Information About a Microarray Experiment (MIAME) - toward standards for microarray data Nature Genetics 2001 29 365 371 11726920
Spellman PT Miller M Stewart J C. T Sarkans U Chervitz S Bernhart D Sherlock G Ball C Lepage M Swiatek M Marks WL Goncalves J Markel S Iordan D Shojatalab M Pizarro A White J Hubley R Deutsch E Senger M Aronow BJ Robinson A Bassett D Stoeckert CJJ Brazma A Design and implementation of microarray gene expression markup language (MAGE-ML) Genome Biology 2002 3 research 0046.1 46.9 12225585
Stoeckert C Parkinson H The MGED Ontology: a framework for describing functional genomics experiments Comparitive and Functional Genomics 2003 4 127 132
Whetzel PL Parkinson H Causton HC Fan L Fostel J Fragoso G Game L Heiskanen M Morrison N Rocca-Serra P Sansone SA Taylor C White J Stoeckert JCJ The MGED Ontology: A resource for semantics-based description of microarray experiments Bioinformatics (under review) 2005
MGED Ontology
FDA Pharmacogenomic Data Submissions
Ball CA Sherlock G Parkinson H Rocca-Sera P Brooksbank C Causton HC Cavalieri D Gaasterland T Hingamp P Holstege F Ringwald M Spellman P Stoeckert CJJ Stewart JE Taylor R Brazma A Quackenbush J An open letter to the scientific journals Science 2002 298 539 12387284
Ball CA Sherlock G Parkinson H Rocca-Sera P Brooksbank C Causton HC Cavalieri D Gaasterland T Hingamp P Holstege F Ringwald M Spellman P Stoeckert CJJ Stewart JE Taylor R Brazma A Quackenbush J An open letter to the scientific journals Bioinformatics 2002 18 1409 12424109
Ball CA Sherlock G Parkinson H Rocca-Sera P Brooksbank C Causton HC Cavalieri D Gaasterland T Hingamp P Holstege F Ringwald M Spellman P Stoeckert CJJ Stewart JE Taylor R Brazma A Quackenbush J An open letter to the scientific journals The Lancet 2002 360 1019
European Bioinformatics Institute, Array Express
Brazma A Parkinson H Sarkans U Shojatalab M Vilo J Abeygunawardena N Holloway E Kapushesky M Kemmeren P Lara GG Oezcimen A Rocca-Serra P Sansone SA ArrayExpress - a public repository for microarray gene expression data at the EBI Nucleic Acids Res 2003 31 68 71 12519949
Rocca-Sera P Brazma A Parkinson H Sarkans U Shojatalab M Contrino S Vilo J Abeygunawardena N Mukherjee G Holloway E Kapushesky M Kemmeren P Lara GG Oezcimen A Sansone SA ArrayExpress: a public database of gene expression data at EBI CR Biol 2003 326 1075 1078
Gene Expression Omnibus
Edgar R Domrachev M Lash AE Gene Expression Omnibus: NCBI gene expression and hybridization array data repository Nucleic Acids Res 2002 30 207 210 11752295
Ikeo K Ishi-i J Tamura T Gojobori T Tateno Y CIBEX: center for information biology gene expression database CR Biol 2003 326 1079 1082
Miyazaki S Sugawara H Ikeo K Gojobori T Tateno Y DDBJ in the stream of various biological data Nucleic Acids Res 2004 32 D31 4 14681352
Young RA Biomedical Discovery with DNA Arrays Cell 2000 102 9 16 10929708
Fellenberg K Hauser NC Brors B Hoheisel JD Vingron M Microarray data warehouse allowing for inclusion of experiment annotations in statistical analysis Bioinformatics 2002 18 423 433 11934741
Manduchi E Grant GR He H Liu J Mailman MD Pizarro AD Whetzel PL Stoeckert CJJ RAD and RAD study annotator: an approach to collection, organization and exchange of all relevant information for high-throughput gene expression studies Bioinformatics 2004 20 452 459 14990441
Tilstone C Vital statistics Nature 2003 424 610 12904757
CSC Microarray Centre, protocols
Oracle Knowledge Base, J Developer
Affymetrix, GDAC Exporter
CSC Microarray Centre, MiMiR download page
CSC Microarray Centre, data annotation tutorial
Lockhart DJ Winzeler EA Genomics, gene expression and DNA arrays Nature 2000 405 827 836 10866209
NCI Metathesaurus
Kemp TJ Causton HC Clerk A Changes in gene expression induced by H2O2 in cardiac myocytes: H2O2 promotes potent and sustained upregulation of p21CIP1/Waf1 Biochem Biophys Res Comm 2003 307 416 421 12859973
ArrayExpress, E-MiMR-12
Oracle, Xquery
Sherlock G Hernandez-Boussard T Kasarskis A Binkley G Matese JC Dwight SS Kaloper M Weng S Jin H Ball CA Eisen MB Spellman PT Brown PO Botstein D Cherry MJ The Stanford Microarray Database NAR 2001 29 152 155 11125075
Gollub J Ball CA Binkley G Demeter J Finkelstein DB Hebert JM Hernandez-Boussard T Jin H Kaloper M Matese JC Schroeder M Brown PO Botstein D Sherlock G The Stanford Microarray Database: data access and quality assessment tools Nucleic Acids Res 2003 31 94 96 12519956
Stanford Microarray Database
Manduchi E Pizarro A Stoeckert C RAD (RNA Abundance Database): an infrastructure for array data analysis Proc SPIE 2001 4266 68 78
RNA Abundance Database
Saal LH Troein C Vallon-Christersson J Gruvberger S Borg Peterson C BioArray Software Environment (BASE): a platform for comprehensive management and analysis of microarray data Genome Biology 2003 3 0003.1 3.6
Maurer M Molidor R Sturn A Hartler J Hackl H Stocker G Prokesch A Scheideler M Trajanoski Z MARS: Microarray analysis, retrieval, and storage system BMC Bioinformatics 2005 6 101 112 15836795
Chen J Zhao P Massaro D Clerch LB Almon RR DuBois DC Jusko WJ Hoffman EP The PEPR GeneChip data warehouse, and implementation of a dynamic time series query tool (SGQT) with graphical interface Nucleic Acids Res 2004 32 D578 581 14681485
Rosetta Resolver
|
16280078
|
PMC1299320
|
CC BY
|
2021-01-04 16:27:46
|
no
|
BMC Bioinformatics. 2005 Nov 9; 6:268
|
utf-8
|
BMC Bioinformatics
| 2,005 |
10.1186/1471-2105-6-268
|
oa_comm
|
==== Front
BMC CancerBMC Cancer1471-2407BioMed Central London 1471-2407-5-1461628198510.1186/1471-2407-5-146Research ArticleIncreased risk of cancer among relatives of patients with lung cancer in China Jin Yongtang [email protected] Yingchun [email protected] Ming [email protected] Saoli [email protected] School of Public Health, Anhui Medical University, Hefei 230032, Anhui province, P R China2 Department of Pharmacology, Anhui Medical University, Hefei 230032, P R China3 Department of Biotechnology, Anhui Medical University, Hefei 230032, P R China2005 11 11 2005 5 146 146 17 7 2005 11 11 2005 Copyright © 2005 Jin et al; licensee BioMed Central Ltd.2005Jin et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Genetic factors were considered as one of the risk factors for lung cancer or other cancers. The aim of this work was to determine whether a genetic predisposition accounts for such familial aggregation of cancer among relatives of lung cancer probands.
Methods
A case-control study was conducted in 800 case families identified by lung cancer patients (probands), and in 800 control families identified by the probands'spouses. The data were analysed with logistic regression analysis model.
Results
The data revealed a significantly greater overall risk of cancer (OR = 1.82, P < 0.01) in the proband group. The relatives of lung cancer probands maintained an increased risk of non-lung cancer (P < 0.05) after adjusting for confounder factors. The crude odds ratio of a proband family having one family member with cancer was 1.67 compared with control families. Proband families were 2.56 times more likely to have two other family members with cancer. For three cancers and four or more cancers, the risk increased to 3.50 and 5.91, respectively. The most striking differences in cancer prevalence between proband and control families were noted for cancer risk among female relatives. The strongest effects were for not only lung cancer in any female relatives (OR 2.17, 95%CI 1.60–3.64) and mothers (OR 2.78, 95%CI 1.23–5.12) and sisters (OR 2.03, 95%CI 1.26–3.97), but also non-lung cancer in females and mothers (OR 2.00, 95%CI 1.26–3.01, and OR 2.34, 95%CI 1.28–4.40, respectively).
Conclusion
These data support the hypothesis of a genetic susceptibility to cancer in families with lung cancer, and the female genetic susceptibility to cancer might be greater than male.
==== Body
Background
Inasmuch as some 22 per cent of all persons die of cancers affecting various anatomic sites, the random occurrence of cancer among several members of a nuclear family would be usual [1]. However, when cancers occur among members in several generations of the same family, chance itself may not be sufficient to account for that occurrence. An increased familial risk for cancer and particularly lung cancer has been demonstrated among relatives of lung cancer patients [2-6]. However, the precise genetic characteristics which influence lung cancer susceptibility are not known. Many studies of families may be used to attempt to disentangle the effects of shared environmental risk factors and genetic factors [7-9].
However, these conclusions do not address the occasionally observed increased frequency of cancers of several sites in combination, such as those of the breast, brain, bone, liver, adrenal cortex, blood, and/or lymphatics[2]. Warthin[10] was the first to describe a large kindred having a remarkable high incidence of cancers at various sites. Pedigree analysis of several large families with multiple cancers across several generations led Lynch and Krush[11] to propose a new genetic entity, later called "cancer family syndrome". In general, those studies showed that the familial occurrence of cancer tends to deviate from the Possion distribution, in that more families than expected by chance have two or more affected, closely related members. However, these high-risk families were selected for study because of their high frequency of cancer cases. Because the laws of probability can not preclude the probability that these cancer families merely represent a chance occurrence, some investigators have questioned genetic susceptibility as a basis for familial aggregation of cancers at various sites.
To explore whether familial aggregation of cancer represents a random or specific event, we conducted a retrospective case-control study of families. The lung cancer patient (proband) was used to identify a case family and the proband's spouse was used to identify a control family in rural regions of Anhui, China.
Methods
Study population
Case-probands were selected if they satisfied a defined eligibility criteria, namely: all farmers who had died of a primary malignancy of the lung (stated as an underlying cause of death in their death records) in three counties in Anhui over a ten-year period (1995–2004). Spouses of probands were included in the control group if these spouses were themselves unaffected with lung cancer and were farmers. The families of probands and controls have lived in rural regions for over 20-year.
The population of the rural regions in Anhui had almost little migration 20 years ago and the cultural patterns and ethnicity have retained definite demographic characteristics. The use of spouse-matched controls can be expected to control for the potential confouning effects of different cultural and residential environment, e.g. diet, drinking water, socioeconomic status and so on.
First-degree relatives (parents, full siblings) of probands were designated as case families. The control families comprised first-degree relatives (parents, full siblings) of the spouses of probands. Thus, in the following sections of this paper, the term "family", when applied to the study population, never includes the proband or spouse.
Data collection
A complete listing of all deaths satisfying the eligibility criteria was obtained from the local Tumor Prevention and Treatment Office of Health Bureau in rural area. Standard demographic characteristics of the probands and the identities of their next-of-kin were obtained from the death records of probands. Local community doctors were recruited to help initiate contact with family members of the case-proband. Provided with the identity of the proband's next-of-kin as well as the usual address of the deceased, the doctors generated a list of addresses of all family members in each pedigree.
Trained interviewers with standard protocols obtained information on each member of the family by face-to-face interviews from (in order of preference) the involved persons (except the proband or other deceased family member), spouse, parent, sibling, or adult offspring. Cancer histories were verified by two methods: 1) a review of death certificates on a sample (90.4 and 88.4 percent, respectively) of relatives of probands and spouses who were reported to have died in rural regions of Anhui and 2) corroborative information from additional family contacts. Because only first-degree relatives (parent and siblings) were included in the study, bias introduced by inability to verify all cancer deaths should be minimal[12].
For the protection of human subjects, all of the subjects in this study signed a consent form according to the guidelines of the World Medical Association Declaration of Helsinki.
Data analysis
Frequency statistics of the study population were computed. Mean age differences between proband and spouse relatives were tested for signficance using two-sided t tests. To determine whether the distribution of relatives was equivalent between study groups, contingency table chisquare tests were used. for each family, design variables were assigned as 1 for the presence and 0 for the absence of each of the following among the proband's or spouse's relatives: one, two, three, four or more cancers. Logistic regression analysis was used to the data to predict whether a family belonged to the case or control group based on these design variables. The resulting regression coefficients (βi)were used to calculate the relative risks of cancer by the formula: odd ratio = eβi, for the ith variables. To determine whether differences in environmental exposure between the case an control families could explain the difference in cancer occurrence, another logistic model was fitted to predict cancer occurrence in each family member based on age, sex, the occupation/industrial exposure pack-years of tobacco exposure and a variable that expressed family membership (case or control). Because the excess risk of cancer in the proband families remained statistically significant, a test of homogeneity was performed to determine whether the study groups differed in their distribution of cancer types. To isolate where these differences occurred, contingency chisquare tests were applied.
Crude odds ratios (OR) were calculated as estimates of the relative risks, and Woolf's method was used to determine 95% confidence intervals (CI)[13]. Maximum likelihood estimates of adjusted ORs were obtained from unconditional logistic regression analysis by the PHREG procedure in SAS software[14].
These variables examined as possible confounders and effect modifiers included: number of first-degree relatives (2–4, 5–7, 8–10, >10); smoking status (never, ex-smoker, current smoker); smoking duration (never, 1–29 yr, 30–45 yr, >45 yr); amount smoked (none, <20 cigarettes per day (cpd), 20–30 cpd, 31–40 cpd, >40 cpd); pack-years (defined as cigarette packs smoked daily multiplied by years of smoking, gram equivalents of leaf tobacco. Assuming 1 g per cigarette) of smoking (none, >0–20, >20–40, >40 pack-years); residence (females only) (selected three counties vs other counties); age (<55 yr, 55–65 yr, >65 yr); ethnicity (Han vs others); sex, passive smoking exposure (ever/never and by total years); education (through high school vs greater than high school and by level: none, primary school, middle school, high school, >high school); martital status (never married, married, and widowed, divorced or separated); high-risk industry/occupation (employed in jobs with exposure to asbestos, benzene, beryllium, bischloromethyl ether, ceramic dust, talc, chemical fertilizers, chromium or chromates, coke oven emissions, dyes, glues, lacquers, fiberglass, cotton dust, insecticides, pesticides, herbicides, fungicides, isopropyl oil, paint sprays, petroleum products, radioactive materials, vinylchloride or explosives); the cumulative exposure to smoky coal use for a given individual was determined by multiplying the annual rate of smoky coal use times the number of years (coal consumption was generally fixed for the households over the life cycle of the family and three exposure categories were formed: >0–70, 70–140, and >140 tons); alcohol consumption (ever/never); vital status; and type of respondent (self/spouse/other); history of COPD (yes/no).
Results
In our study, 800 eligible probands were identified. These families were only included in the data set once (if there are more than one cases in a family, the earliest case diagnosed was selected as the proband). An average of 2.6 and 2.9 interviews (contacts) were made to complete information on each of these proband and spouse families, respectively. The largest proportion of the contacts was through siblings, followed closely by spouses. Less than 17 per cent of the contacts were from adult offspring and surviving parents. The distributions of reported cancer in relatives by source of contacts were not significantly different between proband and spouse families. Information on complete two-generation pedigrees was obtained for 800 case families and 800 control families. The remaining families of probands and spouses were excluded from the analysis because of: inadequate information in names, addresses, incomplete data on each individual in the pedigree (2.8 per cent), or refusals to participate (4.2 per cent). The resulting sample includes 4,453 case family members and 4,378 control family members.
A total of 390 case family members(8.76 per cent) and 212 control family members (4.84 per cent) were reported to have cancer (only primary malignancy counted). 36(9.2 per cent) and 22 (10.4 per cent) were living cancer cases in the case and control groups, respectively. Of the cancer deaths reported in the death certificates, 92.8 per cent were detected by interview. Analogously, 94.2 per cent of the noncancer deaths were reported. There were no significant differences in age at death or age of living family members with cancer between the two study groups.
Male probands almost numbered equally female probands by a ratio of 1.12 to 1.05. The mean family size was similar for both groups, 5.6 including the proband, and 5.5 including the spouse, respectively. Although more of the probands were male and more were older than their spouses, no significant sex or age differences were observed between the two groups of relatives (table 1). The proportional distribution of types of relatives, however, was also non-different between the comparison groups.
Table 1 Characteristics of probands' and spouses' families
Family members and characteristic Proband No. (%) Mean age, yr, of proband Spouse No. (%) Mean age, yr, of spouse
Families 800 800
All relatives 4,453 4,378
Male relatives 2,484(100.0) 2,418(100.0)
Living 1,043(42.0) 56.0 1,064(44.0) 55.3
Dead 1,441(58.0) 49.9 1,354(56.0) 59.4
Female relatives 1,969(100.0) 1,960(100.0)
Living 650(33.0) 57.7 784(40.0) 60.0
Dead 1,319(67.0) 51.0 1,176(60.0) 57.9
Parents 1600(100.0) 1600(100.0)
Livinga 128(8.0) 66.8 192(12.0) 69.0
Dead 1,472(92.0) 60.5 1,408(88.0) 63.1
Siblings 2,853(100.0) 2,778(100.0)
Living 1,883(66.0) 49.0 1,972(71.0) 50.8
Dead 970(34.0) 46.2 806(29.0) 49.0
a P < 0.01, between proportion of relatives in proband and spouse groups who were parents.
Because the proband families tended to be slightly larger than spouse families, adjustment was made for the total number of relatives in the family when determining the familial risk of cancer. Case families were significantly more likely to contain at least one family member with any type of cancer (OR = 1.82, P < 0.01). Each family was further classified according to the total number of cancers present among its menbers (none, one, two, three, four or more). Proband families were 1.67 times more likely to have one family member with cancer and 2.56 times more likely to have two family members with cancer than spouse families(table 2). For three cancers and four or more cancers, the relative risk increased to 3.50 and 5.91, respectively. Each risk other than the lastest estimate was significant at the 0.01 level. The average age of those living family members with cancer, or deceased from cancer, is also presented. The differences are minimal.
Table 2 Familial risk of developing cancer (all sites) among lung cancer case families and control families
Case Control
Cancers (no.) Families (no.) Relativesa (no.) Ageb (M) Families (no.) Relativesa (no.) Ageb (M) ORc (95% CI)
0 518 2369 627 2598
1 209 902 62.2 143 814 61.8 1.67** (1.22–2.01)
2 51 251 61.0 22 112 60.2 2.56** (1.44–4.29)
3 21 95 62.2 7 37 63.1 3.50** (1.36–10.02)
4+ 4 36 62.4 1 17 63.2 5.91 (0.79–132.10)
** p < 0.01
a Excludes probands and spouses.
b Mean age of members living with or deceased from cancer.
c Represents familial risk of case families, not individual risk, after adjusting for family size, total smoky coal exposure, number of person smoking and commune of residence.
To determine whether this excess risk was evident for cancers, we tabulated the total number of cancers in a family including lung cancer (International Classification of Diseases (ICD), 9th Revision, code nos. 162.0–162.9) (table 3). After adjusting for confounder factors, the overall relative risk of one member with cancer in case families (OR 1.54) remained statistically significant. The odds ratios for separate categories also remained significant for families with three members with cancer. There was a trend of risk increased for families with more and more members with cancer. In addition, table 3 significantly shows the increased risk of lung cancer among relatives of lung cancer patients(OR 1.4–5.6, P < 0.05), and the more number of relatives affected by lung cancer, the greater risk of lung cancer among the other relatives of lung cancer patients.
Table 3 Familial risk for developing cancer (non-lung cancer and lung cancer)among lung cancer case families and control families
Case Control
Cancers (n.) Families (no.) Relativesa (no.) Families (no.) Relativesa (no.) ORb (95%CI)
Non-lung cancer
0 657 3,591 720 3,865
1 96 568 63 397 1.54(1.17–2.28)**
2 34 194 12 79 1.67 (0.91–3.91)
3 11 83 4 28 2.11 (1.02–7.98)*
4+ 2 17 1 9 3.13 (0.89–60.66)
Lung cancer
0 658 3629 706 3757
1 113 636 81 528 1.48 (1.11–2.02)*
2 17 103 11 77 2.96 (1.44–6.02)**
3 10 69 2 16 5.67 (1.24–34.89)*
4+ 2 16 0 0 --
* p < 0.05. ** p < 0.01.
a Excludes probands and spouses. b Represents familial risk of case families, not individual risk, adjusted for family size, total smoky coal exposure, number of person smoking and commune of residence.
The results for the analyses that treated lung cancer, non-lung cancer and any cancer among relatives of probands as outcome variables are shown in table 4. Among the 4453 first-degree relatives of the probands, 185 (4.2%) had a reported diagnosis of lung cancer, and 205 (4.6%) had a diagnosis of a non-lung cancer. There were 109 (2.5%) lung cancer cases and 103 (2.4%) non-lung cancer cases among the 4378 first-degree relatives of the spouses. The strongest effects were for not only lung cancer in any female relatives (adjusted OR 2.17, 95%CI 1.60–3.64) and mothers (adjusted OR 2.78, 95%CI 1.23–5.12) and sisters (adjusted OR 2.03, 95%CI 1.26–3.97), but also non-lung cancer in females and mothers (OR 2.00, 95%CI 1.26–3.01, and OR 2.34, 95%CI 1.28–4.40, respectively).
Table 4 Risk of cancer for relatives of lung cancer probands and controls
Relatives Lung cancer ORb (95% CI) Non-lung cancer ORb (95% CI) Any cancer ORb (95% CI)
All relativesa 2.05(1.68–2.53)** 1.41(1.08–2.13)* 1.82(1.52–2.02)**
Female relatives 2.17(1.60–3.64)** 2.00(1.26–3.01)** 2.45(1.84–3.20)**
Male relatives 1.33(1.02–2.02)* 1.03(0.78–1.62) 1.36(1.05–1.68)*
Parents 2.90(1.97–4.32)** 1.52(1.01–2.30)* 2.16(1.62–2.87)**
Fathers 2.17 (1.21–3.86)* 1.30(0.86–2.27) 1.80(1.20–2.69)**
Mothers 2.78(1.23–5.12)** 2.34(1.28–4.40)** 3.12(2.03–4.66)**
Siblings 1.65(1.19–2.18)* 1.31(0.97–2.01)* 1.51(1.28–1.80)**
Brothers 1.21(0.79–1.80) 0.92(0.72–1.68) 1.09(0.82–1.49)
Sisters 2.03(1.26–3.97)** 1.98(1.19–3.21)* 2.13(1.58–3.00)**
* p < 0.05 ** p < 0.01
a Excluded probands and spouses. b OR adjusted for family size, commune of residence, birth oder, total smoky coal exposure, pack-years of smoking, COPD history, education, age, sex and years of cooking.
Discussion
In this study, we attempted to control for as many environmental risk factors as possible for both the probands and their relatives in order to understand better the cause of the observed family aggregation. These data demonstrate an increase risk of non-lung cancer and lung cancer specifically among first-degree relatives of lung cancer probands, and support the hypothesis that genetic factors play a role in the etiology of lung cancer. Tokuhata and Lilienfeld[2] provided the first epidemiologic evidence to support the hypothesis was provided, and reported a significantly increased risk of respiration cancer among relatives of lung cancer patients compared to age-, race-, and sex-matched controls. The risk for cancer in general, however, was not increased. The present authors[3] recently reports similar findings. Lynch et al.[5,15] reported on the familial risk of cancer in relatives of lung cancer probands and probands with other smoking-related cancers (i.e. bladder, pancreas, oral cavity). They found no increased risk for lung cancer when considered separately. However, the risk for cancers of all anatomic sites was significantly increased among the relatives of lung cancer probands (p < 0.001). This increased risk was not evident for relatives of probands with other smoking-associated cancers, nor were the cancers themselves smoking-related. The authors conclude that the data are consistent with the hypothesis of an underlying susceptibility to malignancy in these families. Unlike Lynch et al.[5], the present study again reports an increased risk for lung cancer, and also notes an increased risk for cancer at several anatomic sites in relatives of lung cancer probands. Even after excluding persons with lung cancer and controlling for the assessed environmental exposures, a statistically significant excess risk of cancer was evident.
Our findings also suggest that female relatives (especially mothers) of lung cancer cases are at higher risk for lung cancer than male relatives. Previous studies have indicated that having a first-degree relative with lung cancer was significantly associated with lung cancer among non-smokers[3,6,16-18]. Why the apparently stronger association between lung cancer and lung cancer history in female relatives are unknown. Since certain smoky coal exposure and cooking habits have been implicated as an important risk factor for lung cancer among Chinese women[19]. Genetic factors might influence the susceptibility to environmental carcingens. Excess of 10 carcinogenic chemicals were identified in cooking and smoky coal fume and these shown to interact with one another[20,21] and these possible effects need to be evaluated further.
Findings of familial aggregation and statistical evidence for a major gene [22-24] have led to the search for high-penetrant, rare, single genes for lung cancer and low-penetrant, high-frequency, susceptibility genes for lung cancer[25]. Some of the most widely and recently studied polymorphic loci coding for phase I and II enzymes involved in the activation and conjugation of carcinogen are cytochrome P4501A1 (CYP1A1), glutathione S-transferase M1 (GSTM1), myeloperoxidase (MPO), and NAD(P)H: quinone oxidoreductase (NQO1)[25]. These genes encoding enzymes that are associated with carcinogen metabolism and DNA repair have been the focus of research into possible susceptibility genes for lung cancer and othe cancer. Furthermore, the recent findings show that the homozygous GSTP1 Ile105Val genetype was significantly under-represented in NSCLC compared with controls especially in females, and neither GSTM1 nor MPO genotypes affected the overall risk of NSCLC, and the MPO and CYP1A1 risk genotypes interacted to increase overall risk of NSCLC[26]. Multiple genes of modest effect interact to confer genomic-based susceptibility to lung cancer. It has been hypothesized that differences in susceptibility to carcinogens result from an individual's ability to form genotoxic intermediates, to detoxify these intermediates, and to repair damaged DNA. Polymorphisms in genes coding for the enzymes that drive these processes are likely candidate susceptibility genes. They could not only increase the lung cancer risk but also affect the risk of other cancers. Identifying and testing specific markers using a linkage analysis to confirm the gene involved in the development and progression of lung cancer, particularly to explore the interaction of gene-gene and gene-environment, should be a high future priority.
Abbreviations
ICD, international classification of diseases; COPD, chronic obstructive pulmonary disease; OR, odds ratio; CI, confidence interval; GST, glutathione S-transferase; CYP, cytochrome P450; MPO, myeloperoxidase and NQO1, NAD(P)H: quinone oxidoreductase
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
YJ designed the study and drafted the manuscript. YX carried out the statistical analysis. MX collected the data of lung cancer cases and controls. SX Participated in the design of the study and helped to draft the manuscript. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
We thank those families who took part in the investigation. This work was supported by National Natural Science Foundation of China (NSFC) grants 30471427.
==== Refs
Cairns J The cancer problem Sci Am 1975 233 64 78 1188342
Tokuhata GK Lilienfeld AM Familial aggregation of lung cancer on humans J Natl Cancer Inst 1963 30 289 312 13985327
Ooi WL Elston RC Chen VW Baily-Wilson JE Rothschild H Increased familial risk for lung cancer J Natl Cancer Inst 1986 76 217 222 3456060
Sellers TA Ooi WL Elston RC Chen VW Baily-Wilson JE Rothschild H Increased familial risk of non-lung cancer among relatives of lung cancer patients Am J Epidemiol 1987 126 237 246 3605052
ynch HT Kimberling WJ Markvicks SE Genetics and smoking-associated cancer Cancer 1986 57 1640 1646 3948135
Jonsson S Thorsteinsdottir U Gudbhartsson DF Jonsson HH Kristjansson K Arnason S Guduason V Isaksson HJ Gulcher JR Amundadottir LT Kong A Stefansson K Familial risk of lung carcinoma in the Icelandic population JAMA 2004 292 2977 2983 15613665 10.1001/jama.292.24.2977
Horowitz RI Smaldone LF Viscoli CM An ecogenetics hypothesis for lung cancer in women Arch Intern Med 1988 148 2609 2612 2848471 10.1001/archinte.148.12.2609
Khoury MJ Beaty TH Liang KY Can familial aggregation of disease by explained by familial aggregation of environmental risk factors? Am J Epidemiol 1988 127 674 683 3341366
Ahsan H Thomas DC Lung cancer etiology: independent and joint effects of genetics, tobacco, and arsentic JAMA 2004 29 3026 3029 15613673 10.1001/jama.292.24.3026
Warthin AS Heredity with reference to carcinoma as show by the study of the cases examined in the pathological laboratory of the University Michigan, 1895–1913 Arch Intern Med 1913 12 546 555
Lynch HT Krush AJ Differential diagnosis of the cancer family syndrome Surg Gynecol Obstet 1973 136 221 224 4686764
Schneider NR Chaganti RSK, German J Familial aggregation of cancer Genetics in clinical oncology 1985 New York: Oxford University Press 133 145
Woolf B On estimating the relation between blood group and disease Ann Hum Genet 1954 19 251 254 14388528
SAS Institute. Inc SAS/STAT software: changed and enhancements through release 6.12 1997 Cary, NC: SAS Institute, Inc
Etzel CJ Amos CI spitz MR Risk for smoking-related cancer among relatives of lung cancer patients Cancer Res 2003 63 8531 8535 14679021
Schwartz AG Yang P Swanson GM Familial risk of lung cancer among non-smokers and their relatives Am J Epidemiol 1996 144 554 562 8797515
Wu AH Fontham ET Reynolds P Greenberg RS Buffler P Liff J Boyd P Correa P Family history of cancer and risk of lung cancer among lifetime non-smoking women in the United States Am J Epidemiol 1996 143 535 542 8610670
Shaw GL Falk RT Pickle LW Mason TJ Buffler PA Lung cancer risk associated with cancers in relatives J Clin Epidemiol 1991 44 429 437 2010787 10.1016/0895-4356(91)90082-K
Lan Q Chen W Chen H He XZ Risk factors for lung cancer in non-smokers in Xuanwei County of China Biomed Environ Sci 1993 6 112 118 8397894
Chiang TA Wu PF Ko TC Lee H Genotoxicity of fumes from heated cooking oils produced in Taiwan Environ Res 1999 80 122 126 10092403 10.1006/enrs.1997.3798
Mumford JL Helme CT Lee XM Seidenberg J Nesnow S Mouse skin tumorigenicity studies of indoor coal and wood combustion emissions from homes of residents in Xuan Wei, China with high lung cancer mortality Carcinogenesis (Lond) 1990 11 397 403 2311182
Jin YT Xu YC Yang RD Xu CW He XZ Familial aggregation of lung cancer in a high incidence area in China Brit J cancer 2005 92 1321 1325 15756270 10.1038/sj.bjc.6602465
Sellers TA Ooi WL Elston RC Chen VW Increased familial risk for non-lung cancer among relatives of lung cancer patients Am J Epidemiol 1987 126 237 246 3605052
Wu PF Lee CH Wang MJ Goggins WB Chiang TA Huang MS Ko YC Cancer aggregation and complex segregation analysis of families with female non-smoking lung cancer probands in Taiwan European Journal of Cancer 2004 40 260 266 14728941 10.1016/j.ejca.2003.08.021
Schwartz G Genetic predisposition to lung cancer Chest 2004 125 86S 89S 15136430 10.1378/chest.125.5_suppl.86S
Larsen JE Colosimo ML Yang IA Bowman R Zimmerman PV Fong KM CYP1A1 Ile462 and MPO G-463A interact to increase risk of adenocarcinoma but not squamous cell carcinoma of the lung Carcinogenesis 2005, Sep 29
|
16281985
|
PMC1299321
|
CC BY
|
2021-01-04 16:39:11
|
no
|
BMC Cancer. 2005 Nov 11; 5:146
|
utf-8
|
BMC Cancer
| 2,005 |
10.1186/1471-2407-5-146
|
oa_comm
|
==== Front
BMC Cardiovasc DisordBMC Cardiovascular Disorders1471-2261BioMed Central London 1471-2261-5-331628393010.1186/1471-2261-5-33Study ProtocolCardiovascular disease, risk factors and heart rate variability in the elderly general population: Design and objectives of the CARdiovascular disease, Living and Ageing in Halle (CARLA) Study Greiser Karin H [email protected] Alexander [email protected] Barbara [email protected] Jan A [email protected] Cees A [email protected] Oliver [email protected] Karl [email protected] Johannes [email protected] Institute of Medical Epidemiology, Biostatistics and Informatics, Martin-Luther-University Halle-Wittenberg, Magdeburger Str. 27, 06097 Halle (Saale), Germany2 Department of Medical Informatics, Erasmus Medical Center Rotterdam, PO Box 1738, Rotterdam, The Netherlands3 Department of Cardiology, Leiden University Medical Center, PO Box 9600, 2300 RC Leiden, The Netherlands4 Department of Medicine III, Martin-Luther-University Halle-Wittenberg, Ernst-Grube-Str. 40, 06097 Halle (Saale), Germany2005 11 11 2005 5 33 33 2 10 2005 11 11 2005 Copyright © 2005 Greiser et al; licensee BioMed Central Ltd.2005Greiser et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The increasing burden of cardiovascular diseases (CVD) in the ageing population of industrialized nations requires an intensive search for means of reducing this epidemic. In order to improve prevention, detection, therapy and prognosis of cardiovascular diseases on the population level in Eastern Germany, it is necessary to examine reasons for the East-West gradient of CVD morbidity and mortality, potential causal mechanisms and prognostic factors in the elderly.
Psychosocial and nutritional factors have previously been discussed as possible causes for the unexplained part of the East-West gradient. A reduced heart rate variability appears to be associated with cardiovascular disease as well as with psychosocial and other cardiovascular risk factors and decreases with age. Nevertheless, there is a lack of population-based data to examine the role of heart rate variability and its interaction with psychosocial and nutritional factors regarding the effect on cardiovascular disease in the ageing population. There also is a paucity of epidemiological data describing the health situation in Eastern Germany. Therefore, we conduct a population-based study to examine the distribution of CVD, heart rate variability and CVD risk factors and their associations in an elderly East German population. This paper describes the design and objectives of the CARLA Study.
Methods/design
For this study, a random sample of 45–80 year-old inhabitants of the city of Halle (Saale) in Eastern Germany was drawn from the population registry. By the end of the baseline examination (2002–2005), 1750 study participants will have been examined. A multi-step recruitment strategy aims at achieving a 70 % response rate.
Detailed information is collected on own and family medical history, socioeconomic, psychosocial, behavioural and biomedical factors. Medical examinations include anthropometric measures, blood pressure of arm and ankle, a 10-second and a 20-minute electrocardiogram, a general physical examination, an echocardiogram, and laboratory analyses of venous blood samples. On 200 participants, a 24-hour electrocardiogram is recorded. A detailed system of quality control ensures high data quality. A follow-up examination is planned.
Discussion
This study will help to elucidate pathways to CVD involving autonomic dysfunction and lifestyle factors which might be responsible for the CVD epidemic in some populations.
==== Body
Background
Cardiovascular diseases (CVD) are the leading cause of death and morbidity in industrialized nations, accounting for about 50% of all deaths [1]. In old age, both incidence and prevalence of CVD are increasing. Although the mortality of coronary heart disease has decreased since the 1970s, the prevalence, incidence and mortality of chronic heart failure has persisted or even increased [2-5]. In the ageing population, chronic heart failure (CHF) is an increasing public health problem and poses a mounting economic burden on societies and their health care systems due to its association with frequent hospitalizations and the need for long-term pharmacological treatment [2,6-9].
Moreover, in spite of the general decline of cardiovascular mortality over the past decades, an increasing East-West gradient within Europe has been observed with decreasing life-expectancy and higher rates of cardiovascular morbidity and mortality in Eastern European countries which seems to arise from premature CVD [10,11]. This discrepancy has grown since the political changes and socioeconomic discontinuities in Eastern Europe at the beginning of the 1990s and cannot be explained entirely by the classical risk factors for cardiovascular diseases alone [12]. Even within Germany, an East-West gradient of CVD mortality is still present after re-unification [13]. However, there is a lack of population-based data on CVD morbidity and risk factors to further analyse the extent, cause and potential for reduction of the increased CVD burden in the East. Among the newer risk factors discussed as potential explanations for the East-West gradient are nutritional factors, alcohol and psychosocial factors [10,14,15]. Psychosocial factors such as depression, social isolation, job strain and hostility are associated with higher CVD risk [16-19] and are differentially distributed by socioeconomic status [20-23].
Autonomic dysfunction as indicated by a reduced heart rate variability (HRV) is increasing with age [24,25] and has been observed to be associated with an increased risk for incident coronary heart disease (CHD), CVD mortality [26-30] and worse prognosis in patients with CHD or heart failure [31-34]. However, there is growing evidence that adverse psychosocial factors might also be associated with a reduced HRV and other measures of sympathovagal imbalance [35-38]. A reduced heart rate variability appears to integrate negative factors such as disease and stress. Thus, autonomic dysfunction, of which HRV is a marker, might play an important role in mediating the effect of social inequality, psychosocial factors and other cardiovascular risk factors on CVD causation and prognosis [39-42].
For inflammatory factors such as C-reactive protein (CRP) and cytokines (tumour necrosis factors, interleukin-6), higher levels have been observed in elderly subjects and in subjects at higher risk of chronic heart failure (CHF) [43,44]. They have also been associated with a reduced HRV [45]. However, the role of inflammatory factors and their interaction with HRV in the pathogenesis of heart failure is still not well understood. There is a lack of population-based studies to investigate the potential mechanisms involving HRV, inflammatory factors and psychosocial factors underlying causation and prognosis of CHF.
The examination of heart rate variability in a general population might provide valuable information to elucidate some of the pathways relevant for the development and prognosis of CVD. It might also help to explain the mechanisms underlying the observed associations of psychosocial factors with CVD. Moreover, unfavourable psychosocial factors and a reduced heart rate variability could be among the factors responsible for the higher CVD mortality in Eastern European and Eastern German populations as compared to their Western counterparts. The examination of reasons for regional variation such as the East-West gradient of CVD morbidity and mortality as well as of potential mechanisms of causation and of prognostic factors in the elderly could lead to advances in the prevention, detection, therapy and prognosis of cardiovascular diseases on the population level.
Furthermore, the role of HRV as a potential additional screening tool for the identification of subjects at high risk of acquiring cardiovascular diseases or of suffering fatal adverse events should be explored. However, there is a lack of population-based data regarding the distribution of HRV and its interaction with other CVD risk factors in the general ageing population. The recording of 24-hour holter electrocardiograms (ECG) or ECGs of at least 2 hours duration is still the recommended standard for HRV analysis in clinical practice [46]. While 24-hour ECGs record the dynamic component of HRV derived from fluctuations in daily activities and thus contain other information than short-term ECGs, a great advantage of short ECG recordings in resting state is that they can be highly standardized and are easier to apply in larger populations. For screening purposes in a general population, the recording of long-term ECGs is impractical.
Although short-term and long-term ECGs measure different components of HRV, one study in healthy young men showed a high predictive value of short-term ECGs for long-term ECGs [47]. Unfortunately, there is a lack of studies assessing the validity and predictive value of short-term ECGs for the determination of HRV as compared to HRV measures derived from 24-hour ECGs in the general, elderly population. Therefore, it is important to examine whether ECGs of shorter recording duration (e. g., 20 minutes) are a valid basis for the identification of subjects with reduced HRV in the general population.
This paper presents the design and objectives of the CARdiovascular disease, Living and Ageing in Halle (CARLA) Study.
The aims of this study are:
1. to examine the distribution of different parameters of heart rate variability and the prevalence of a reduced HRV in a representative sample of an elderly East German population;
2. to examine the prevalence of cardiovascular diseases and their association with HRV, psychosocial and socioeconomic factors, and inflammatory and classical cardiovascular risk factors in an ageing general population;
3. to identify factors responsible for the epidemic of chronic heart failure in the ageing population, and to identify potential areas of prevention;
4. to elucidate reasons for regional variations in CVD morbidity and mortality; and
5. to assess the potential of HRV derived from short-term ECGs as a screening tool in the general population for the detection of subjects at high risk of CHD or CHF and for risk stratification in chronic heart failure or other cardiovascular diseases.
A planned follow-up study will allow the determination of incident events.
Methods/design
The design of the CARLA study is a population-based epidemiologic cross-sectional study with the aim of a prospective follow-up of the examined study subjects, thus leading to a cohort study. The study was approved by the Ethics Committee of the Medical Faculty of the Martin-Luther-University Halle-Wittenberg and by the State Data Privacy Commissioner of Saxony-Anhalt. The study is conducted by the Institute of Medical Epidemiology, Biostatistics and Informatics in cooperation with the Department of Internal Medicine III (Cardiology, Angiology and Medical Intensive Care) at the Martin-Luther-University Halle-Wittenberg.
Study population, sampling and recruitment procedure
The base population for the study participants are male and female inhabitants of German nationality of the city of Halle (Saale) in Saxony-Anhalt, Eastern Germany, aged 45 to 80 years. The city of Halle has approximately 240 000 inhabitants and is the largest city in Saxony-Anhalt. A random sample of 5000 men and women aged 45 to 80 years at the time of the sampling (July 2002) was drawn from the population registry of the city of Halle.
The sampling was done within the age strata 45–49, 50–54, 55–59, 60–64, 65–69, 70–74 and 75–80 years. In all but the last stratum, equal numbers of men and women were drawn (322 women and 323 men per stratum). The age stratum 75–80 years was oversampled with twice as many inhabitants drawn into this sample than for each of the other strata (625 women and 625 men). This was done in order to account for lower expected response rates in the oldest group.
The recruitment of study subjects began in December 2002 and is aimed to be completed by the end of the year 2005. When completed, the final study population will comprise 1750 persons. Up to July 2005, about 1500 subjects were examined. Assuming comparable response rates within all age-sex strata except the oldest one, we expect to recruit about 120 men and 120 women in each 5-year-age group from 45 to 74 years, and 155 men and women each in the age-group 75 to 80 years. An overall response rate of 70 % is aimed at.
The recruitment of study subjects has been done by inviting consecutive waves of random sub-samples of the original population sample. Therefore, not all persons originally drawn from the population registry have to be invited in order to obtain a representative sample of the Halle population aged 45 to 80 years.
The method of recruitment applied in the CARLA study was selected after discussion with experts about the most efficient strategies to achieve the desired response rate. The recruitment strategy includes at least two written invitations and active contact attempts by the study personnel via telephone or home visit if the eligible person does not respond to the written invitation. The mailed invitation contains a letter of invitation, a detailed description of the aims and examination procedures of the study, and a copy of the approval letter of the Data Privacy Commissioner for Saxony-Anhalt. Different recruitment strategies are followed for subjects for whom a home telephone number can be identified from the official telephone directories and for those with an ex-directory or no telephone number:
Persons with known telephone numbers are identified and receive a letter of invitation announcing a subsequent phone call by the study personnel. If the person does not respond actively to the invitation letter, study personnel make at least 10 attempts to establish telephone contact at different times of the day and on different days. If no contact can be made over at least six weeks, a reminder invitation is mailed to the subject, offering the option of replying with a prepared reply-paid card indicating convenient examination dates. If the second invitation does not lead to a contact with the subject or to verification of his response state (e. g. moved with unknown address, too ill to participate in the study, deceased, or unwilling to participate in the study), a home visit is performed by study personnel. Several attempts are made to localize the study subject or to retrieve information from relatives or neighbours regarding the subject's correct address of residence or telephone number, or their willingness or ability to participate.
For those subjects without a registered telephone number, two invitation letters with reply-paid cards are sent before entering the home visit phase.
All direct contact attempts (phone or home visit) are made by trained and certified study personnel and are documented in detail. At each home visit which does not result in contact with the invited subject or another member of the household, the home visitor leaves a pre-printed note indicating that a visit has been made and indicating how contact can be made with the researchers. Figure 1 shows the recruitment scheme of the CARLA Study.
Figure 1 Recruitment scheme for CARLA baseline examination.
A letter confirming the date of examination is sent to subjects who agreed to participate in the study and contains a consent form, a map of the study centre and a self-administered questionnaire.
Data collection – interview, questionnaires and examination procedures
In order to provide a basis for pooled data analyses and to permit valid comparisons with other study regions and populations, the CARLA Study deliberately uses highly standardized and validated instruments of data collection which have been applied in several completed or ongoing national and international cardiovascular disease epidemiologic studies. Questionnaire items and examination procedures were selected and adapted from i.) the KORA Study Augsburg in Bavaria, Southern Germany (the continuation of the MONICA Surveys in Augsburg) [48]; ii.) the SHIP Study Greifswald in Pomerania, North-Eastern Germany [49]; iii.) the EPIC Potsdam Study, Eastern Germany [50]; iv.) the Rotterdam Study, The Netherlands [51]; and v.) the HAPIEE Study which includes study regions in the Czech Republic, Poland, Russia and Lithuania [52].
After selection of interview items and examination procedures, a pretest was performed to test instruments and logistics of data collection. Slight modifications were made before recruitment of the study subjects started.
The data collection consists of a standardized, computer-assisted personal face-to-face interview, a self-administered questionnaire, a medical examination by a trained and certified study nurse, and a physical examination and a transthoracic echocardiogram performed by a physician who has been specifically certified for this study. The examination takes place at the CARLA study centre in the University Hospital of the Martin-Luther-University Halle-Wittenberg. The average duration is 3.5 hours per subject.
Interview
The computer-assisted interview was programmed with the DAIMON interview programme which was also used in the KORA 2000 Survey Augsburg [53]. The interview collects information on sociodemographic and socioeconomic data, medical history and cardiovascular risk factors. In detail, it includes questions on household income, educational level and occupation of the interviewee, and his or her parents and partner (where appropriate), material circumstances, psychosocial factors such as social support, unemployment and job security, changes in socioeconomic factors and social relations since the German re-unification in 1990, utilization of health care services, family medical history of chronic diseases, and history of any physician-diagnosed cardiovascular diseases (coronary heart disease, stroke, arterial hypertension), diabetes, hypercholesterolemia, thyroid disease, osteoporosis, chronic bronchitis, rheumatoid arthritis and cancer. It furthermore contains the Rose questionnaire on angina pectoris and intermittent claudication [54] and information on lifestyle-factors such as smoking, diet, alcohol consumption and physical activity. The interview lasts approximately one hour. Details of the sources of the interview modules are listed in table 1. Information on the use of medication during the seven days preceding the examination is collected by the study nurse with the computer-based IDOM programme [55]. It integrates the GKV medication database, the official medication database of the Wissenschaftliches Institut der Ortskrankenkassen (WIdO, Scientific Institute of the Local Sickness Funds) which is continuously updated and allows an automated search for individual drugs by PZN number (central pharmanumber) and direct retrieval of ATC codes [56].
Table 1 Components and sources of the interview used in the baseline survey of the CARLA Study
Topics of interview module Source (reference, original study)
Sociodemographic factors:
1. Subject's own education, occupational status, net household income
2. Parental education and occupational status
3. Partner's education and occupational status 1. Demographic standards for Germany [81]
2. adapted according to [81]
3. adapted according to [81]
Utilization of medical services Modified from SHIP [49,82]
Chronic diseases: own medical history
1. cardiovascular diseases: coronary heart disease, stroke, hypertension
2. chest pain
3. intermittent claudication
4. hypercholesterolemia
5. diabetes
6. thyroid disease
7. osteoporosis
8. rheumatoid arthritis
9. chronic diseases
10. cancer
Chronic diseases: family history
1. myocardial infarction, stroke, hypertension
2. diabetes
3. cancer
Menopausal state and use of hormone replacement therapy All adopted from SHIP [49,82] and KORA/MONICA [48,83-85];
Rose questionnaire for chest pain and intermittent claudication [54]
Medication use during the preceding 7 days IDOM software developed by GSF, based on the GKV medication database for Germany, used in KORA study [55,56]
Health related behaviour/lifestyle factors
1. diet/nutritional habits – short qualitative food frequency list
2. alcohol use
3. physical activity
4. smoking 1. adopted from KORA/MONICA [83,86]
2. adopted from HAPIEE [52]
3. Baecke questionnaire [87], also used in SHIP and ARIC
4. adapted and shortened from smoking questionnaire developed by W. Ahrens, BIPS
Social support Modified version of Berkman-questionnaire (translation by J. Siegrist used in KORA/MONICA) [48,84,88,89]
Unemployment, job insecurity Adopted from SHIP [49,82] and HAPIEE [52]
Material circumstances Adopted from HAPIEE [52]
ARIC = Atherosclerosis Risk in Communities; BIPS = Bremen Institute for Prevention Research and Social Medicine; CES-D = Center of Epidemiological Studies Depression Scale; EPIC = European Prospective Investigation into Cancer and Nutrition; HAPIEE = Health, Alcohol and Psychosocial Factors in Eastern Europe; KORA = Cooperative Health Research in the Augsburg Region; MONICA = Monitoring Trends and Determinants in Cardiovascular Disease; SF 12 = Short Form Health Survey Questionnaire; SHIP = Study of Health in Pomerania
Self-administered questionnaire
The self-administered questionnaire contains the food-frequency questionnaire (FFQ) also used in the EPIC Potsdam Study [50] follow-up, quantitative questions about alcohol drinking patterns aimed at obtaining information on binge drinking, questions about health beliefs, social networks, the CES-D questionnaire on depression, the SF-12 questionnaire on self-rated health and health-related quality of life, the effort-reward imbalance questionnaire developed by Siegrist [57,58], and questions about security in the neighbourhood environment before and after the re-unification in 1990. The EPIC FFQ is usually completed at home after the examination date and sent back via reply-paid envelope.
Details on the sources of the questionnaire modules are listed in table 2.
Table 2 Components and sources of the self-administered questionnaire used in the baseline survey of the CARLA Study
Topics of questionnaire modules Source (reference, original study)
Health related behaviour/lifestyle factors
1. diet/nutritional habits – quantitative food frequency questionnaire (FFQ)
2. alcohol use – binge drinking 1. adopted from EPIC [50]
2. adopted from HAPIEE [52]
Social networks Modified version of Berkman-questionnaire (translation by J. Siegrist used in KORA/MONICA) [48,84,88,89]
Perceived security in neighbourhood environments before and after German re-unification in 1990 adopted and translated from HAPIEE [52,90-92]; similar questions in SHIP [82]
Health beliefs Adopted from HAPIEE [52]
Health-related quality of life Social functioning questionnaire SF 12 [93]
Job strain, effort-reward imbalance Effort-reward imbalance questionnaire by J. Siegrist [57,58]
Depression scale German translation of the CES-D depression scale by Kohlmann & Gerbershagen [94-96]
Abbreviations: same as in table 1
Examination procedures
The medical examination includes the measurement of anthropometric parameters and arterial blood pressure, the determination of the ankle-arm index of systolic blood pressure, the recording of a 10-second- and a 20-minute electrocardiogram, a trans-thoracic echocardiogram, and the drawing of a venous blood sample.
At the beginning of the examination, the subject is seated comfortably in a chair and asked to assume the position required later for the correct measurement of sitting blood pressure. Then, the medication currently taken by the study subject is recorded using the IDOM programme (see table 1). After a resting period of at least five minutes, the measurement of systolic and diastolic blood pressure is started. Blood pressure is measured with the OMRON HEM-705CP automated oscillometric blood pressure device [59] according to the procedure employed in the SHIP and KORA/MONICA Study [48,49,55]. The OMRON HEM-705CP device meets the criteria defined for the use in clinical trials by the Association for the Advancement of Medical Instrumentation and the British Hypertension Society. The size of the cuff is selected according to the arm circumference (circumference <32 cm: normal adult cuff, circumference 32 – <42 cm: large adult cuff). Three measurements are performed on the left arm with a three-minute delay between each pair of measurements. Heart rate is counted manually during the resting period. Data entry of the measurements is done immediately using the DAIMON screen developed by the GSF which includes a timer for the three-minute delay between the blood pressure measurements [53].
The anthropometric measurements follow the procedures used in the MONICA/KORA and SHIP study [48,49,55]. Weight and height are measured with the SECA 701 digital scale and the SECA 220 height measuring system. Waist and hip circumference are measured using a flexible tape, with the study subject standing in front of a full-size mirror which allows checking the horizontal position of the tape. Weight is recorded with a precision of 100 g, and height, waist and hip circumference to the nearest 0.1 cm.
The individual is then asked to lie down and rest for at least five minutes before the measurement of supine systolic blood pressure at arm and ankle is commenced for the determination of the ankle-arm index as indicator of peripheral arterial disease (PAD). The protocol for the measurement of ankle-arm index using the OMRON HEM-705CP both at arm and leg was developed by the CARLA Study. First, a simultaneous measurement of blood pressure (BP) at both arms is performed. For the BP measurements used for the calculation of ankle-arm index, the OMRON HEM-705CP device remains on the arm with the higher systolic blood pressure. The circumference of both calves is determined at the midpoint of the BP cuff to select the adequate cuff size. The cuff is positioned approximately 5 cm above the inner ankle over the posterior tibial artery in contour wrap technique [60]. Measurement of BP is started simultaneously on arm and ankle, using two OMRON HEM-705CP devices at the same time. First, two measurements are performed at the right ankle, followed by two measurements at the left ankle. Between each pair of measurements, there is a one-minute delay.
Following the supine BP measurement two resting electrocardiograms (ECG) are recorded: one 10-second and one 20-minute 12-lead ECG. The study subject remains in supine position and is not allowed to raise from the beginning of the supine BP measurements until completion of the ECGs. The attachment of the electrodes follows the standard procedures for measuring and marking of the electrode positions using a DAL-square as described in the ARIC Manual 5 Electrocardiography [61] and adopted by the KORA/MONICA Study [55]. The ECGs are recorded using a Cardio Control Medical Diagnostic Workstation 1.3.1 with a Cardio Perfect MD Recorder with a sampling rate of 600 Hz (Welch Allyn Cardio Control, Delft, NL) and stored digitally. The study subject is asked to remain in supine position and not to speak throughout the ECG recording. The 10-second ECG is recorded first, and a printout is provided for interpretation by the study physician.
It is required that the study subject has remained in supine resting position for at least 20 minutes before the recording of the 20-minute ECG is started. The baroreceptors which sense the blood pressure in the aorta and in the carotid sinuses are instrumental in the genesis of HRV. The 20-minute supine resting period ensures that the baroreceptors have adapted to the change in blood pressure due to assuming the supine position. Throughout the 20-minute ECG, the subject is asked to breathe at a frequency of 15/min (0.25 Hz). For guidance of the exact respiratory rhythm at 0.25 Hz, the Leiden Respiratory Metronome, a visual metronome developed by H. v. d. Vooren and M. Santunione (©Leiden University Medical Center, Cardiology Department, used with permission by C. A. Swenne) is displayed on a computer screen easily visible for the subject. The purpose of the metronome guided respiration is to standardize the ECG recording as much as possible with respect to the influence of the respiratory rate on the determination of spectral parameters of the HRV. Thus, by ensuring a standard supine resting period of 20 minutes preceding the ECG recording and an equal respiratory rate for all subjects, the major physiological mechanisms influencing short term HRV are controlled for, and any differences in HRV found between subjects or groups of subjects can be attributed to other factors that are characteristic to the study subject (e. g. pathological processes, or influence of other risk factors on HRV).
After the ECG recording, a non-fasting venous blood sample is drawn under standardized conditions according to the KORA/MONICA protocol [48,55,62] with the subject remaining supine.
A transthoracic echocardiogram is performed on the GE Vingmed Ultrasound Vivid Five System or on the GE Vingmed System Five Performance with a FPA 2.5 MHz probe and Echo Pac Software version 6.3 (GE Medical Systems Ultrasound, GE Ultraschall Deutschland, Solingen) by a certified study physician following the protocol of the SHIP Study [49,63]. It collects information on parameters derived from M-Mode and Doppler standard echocardiographic techniques. Parameters to be studied include measurements required for the determination of left ventricular mass and systolic and diastolic function. Images and measurements of the echocardiographic examination are saved on optical disc and video to permit offline reading. The examination is concluded with an explication of the immediately available examination results (e. g., elevated blood pressure values, major electrocardiographic or echocardiographic abnormalities) for the study subject by the study physician. Table 3 gives an overview of the examination components and equipment.
Table 3 Examination components and equipment of the CARLA Study
Examination components Parameters Instruments
Anthropometry Body weight, body height, waist- and hip circumference, body mass index Digital scales (SECA 701), body height measuring system (SECA 220), tape measure
Blood pressure Systolic and diastolic blood pressure at the arm and the ankle, heart rate OMRON HEM-705CP automated oscillometric blood pressure measurement device
Electrocardiogram Minnesota-Code (MEANS), HRV parameters, general de- and repolarisation parameters in the ECG (LEADS) 1. Welch Allyn Cardio Control Diagnostic Medical Workstation, MD Recorder, 600 Hz sampling rate (10-sec/20-min ECG)
2. mtm Multitechmed DMS 300-9 holter ECG recorder, 1024 Hz sampling rate
Echocardiogram Measures of systolic and diastolic function, heart valves, ventricular dimensions GE Vivid Five/System Five with FPA 2.5 MHz probe
For a random sub-sample of 200 subjects, a 24-hour ECG is recorded with the DMS 300-9 holter ECG recording system with a sampling rate of 1024 Hz (mtm Multitechmed, Hünfelden, Germany).
Laboratory analyses
The non-fasting venous blood samples include the collection of serum, EDTA plasma, citrate plasma and EDTA full blood. All serum and plasma samples are centrifuged at 4°C in-house by specially trained study laboratory personnel. A small part of the serum and EDTA plasma is sent for analysis of total, HDL and LDL cholesterol, triglycerides, glucose, haemoglobin A1c (HbA1c), creatinine, high-sensitive C-reactive Protein (CRP) and n-terminal brain natriuretic peptide (NT-proBNP) to a single central laboratory at the University of Leipzig. All laboratory analyses – except for the determination of cytokines – are undertaken by the Institute of Laboratory Medicine, Clinical Chemistry and Molecular Diagnostics (ILM) at the Leipzig University Clinics. The laboratory has been accredited according to the accreditation norms ISO 15180 and ISO 17025. The cytokines interleukin-1 and 6 (IL-1, IL-6) and tumour necrosis factor-alpha (TNF-α) are determined at the research laboratory of the Department of Medicine III at the Martin-Luther-University Halle-Wittenberg. The major part of the specimens is stored at -80°C for future analysis. Citrate plasma is stabilized with 10% metaphosphoric acid (0.1 ml plasma : 0.9 ml MPA) before freezing for subsequent analyses of ascorbic acid (vitamine C). Serum total cholesterol is analyzed with the enzymatic colorimetric CHOD-PAP method at the Modular system (Roche Diagnostics, Mannheim, Germany) [64]. A homogenous enzymatic colorimetric test is used for the determination of serum HDL and LDL cholesterol at the Modular system [64,65]. Serum triglycerides are measured applying the colorimetric enzymatic GPO-PAP assay at the Modular system [64]. Casual serum glucose is measured on the Modular system with the hexokinase method. For the analysis of HbA1c in EDTA anticoagulated blood, the High Performance Liquid Chromatography (HPLC) method is used with the Variant II system (BIO-RAD, Munich, Germany). Creatinine is measured colorimetrically enzymatically on the Modular system [64]. High-sensitive C-reactive proteine (hsCRP) is determined with a turbidimetric immunologic reaction assay enhanced by the addition of latex on the Modular system [64,66]. NT-proBNP is measured using an electro-chemo-illuminescence assay by Roche Diagnostics on the Elecsys system (Roche Diagnostics, Mannheim, Germany) [67,68]. IL-1, IL-6 and TNF-α are measured using ELISA-techniques.
Genetic analyses including the haplotype mapping of specific gene variants (SNPs) using a pathway-oriented approach are planned. SNPs of the following signal and metabolic paths important for ageing will be analysed using blood stored at -80°C: 1.) the TNF-α/NF-κB pathway relevant for inflammatory factors and immune senescence; 2.) the IGF-1/PI3K/Akt longevity pathway; 3.) candidate genes for ageing (RAGE, ApoE, Klotho, interleukin-6 promotor polymorphism). Genes showing a positive association with a reduced HRV or a reduced cardiac function indicating cardiac ageing will be examined regarding their specific function and regulation in the context of cardiac ageing.
For a random subsample of 400 subjects, further laboratory parameters from frozen samples will be analysed, e. g. lipoprotein (a), vitamins, folic acid, homocysteine and thyroid hormones.
ECG analyses
All 10-second ECGs are processed by the Modular ECG Analysis System (MEANS) [69] to obtain Minnesota Codes [70]. The MEANS program has been extensively validated [69,71-73]. MEANS also processes the 20-minute ECGs to obtain the locations and types of the QRS complexes. This information is then used to analyse HRV computing standard time domain and frequency domain parameters of HRV. HRV is computed according to the procedure used in previous studies. The method has been described elsewhere [47,74,75]. Briefly, artefacts and ectopic beats are replaced by interpolated normal sinus beats to a maximum of ten percent. For each ECG, the percentage of replaced beats is calculated. ECGs which display non-sinus rhythm, like atrial fibrillation, or where the majority of QRS complexes is paced by an artificial pacemaker, are excluded from HRV analysis.
Then, time domain parameters of HRV are calculated. The calculated time domain HRV parameters include the standard deviation of normal intervals (SDNN), the mean absolute successive normal interval differences, the root mean square successive normal interval differences (rMSSD), and the percentage of successive normal interval differences >50 ms (pNN50).
Prior to spectral analysis of the ECG, the tachogram has to undergo further processing steps. The further pre-analytic treatment of the inter-beat-interval (IBI) tachogram, which is the graphical presentation of RR intervals over time, includes an adjustment for linear trends, tachogram tapering and zero padding. A test of stationarity is performed, and non-stationary recordings are excluded from analysis. ECGs with extreme HRV values or high in-stationarity scores are visually checked to identify possible sources of error or explanations for the extreme values. If only part of the ECG is responsible for the lack of overall stationarity, stationary sub-sections of the 20-minute ECG are identified and selected for repeated HRV analysis.
For spectral density estimation of frequency domain variables, a fast Fourier transformation is employed to calculate very low frequency power (VLF power), low frequency power (LF power), high frequency power (HF power) and total power (TP).
Those HRV parameters are also calculated from 24-hr ECGs. The ranking of the subject regarding the distribution of HRV parameters in the study population and the diagnosis of a reduced HRV are compared between results derived from 20-minute ECGs and from 24-hour ECGs.
The addition of further ECG parameters – such as parameters describing T-wave complexity [76], the ventricular gradient, and the spatial angle between the QRS and T axes [77] – characterizing relevant physiological and pathophysiological processes which might improve the understanding of cardiovascular disease mechanisms – have been incorporated into the MEANS programme (MI-EUR) and into the LEADS programme (LUMC) [78] and are available for future analysis.
Echocardiographic analyses
The transthoracic echocardiographic examinations are stored as digital files in order to permit off-line reading of the parameters independent of the examination procedure itself. However, for immediate documentation and interpretation by the study physician, online reading is performed during the examination. The analysis includes parameters of left ventricular dimension (left ventricular mass) and of systolic and diastolic function derived from M-Mode and Doppler echocardiographic measurements.
Quality control – training and certification of study personnel and data quality management
In order to ensure a highly standardized data collection, all study personnel is specifically trained and certified for the study procedures. Interviewers and the study nurse were trained and certified in cooperation with the KORA study centre Augsburg before commencement of data collection in the study. Apart from certification examinations, a high level of standardization and examination quality is ensured by repeated supervisions of the study nurse and interviewers (by the principal investigator) during the examination or interview of study subjects.
For means of quality control, all interviews are digitally recorded, and a randomly selected sample of 10% of all interview recordings is monitored by comparing audiorecording with the data entry. In addition, interviews found to contain implausible answers during the computerized plausibility checks are monitored using the audiorecording.
The CARLA study physicians participate in the observer and reader certification for the echocardiographic examination held at the SHIP study centre Greifswald for echocardiographers of the SHIP and KORA study.
All paper documentation is double entered in order to minimize errors due to data entry. Visual and computerized plausibility checks are performed to detect possible data entry errors of paper documentation, examination procedures or interview. (The plausibility checks applied to ECG data have been described briefly above.)
Statistical analysis
The statistical analysis includes the calculation of frequency distributions including means and medians of risk factors and HRV parameters by age, sex and cardiovascular disease status, the calculation of the prevalence of adverse risk factor levels and cardiovascular diseases, the calculation of correlation coefficients, and linear and logistic regression analyses. Differences between groups are tested using the t-test, the F-test or a test for trend. Standard errors, standard deviations and 95% confidence intervals are calculated to evaluate precision of the estimates. All statistical analyses are performed with SAS, Version 9 (SAS Institutes, Cary, NC). To assess validity of the 20-minute ECG compared to the gold standard (24-hour ECG), standard measures for diagnostic tests (sensitivity, specificity, positive and negative predictive value) are calculated.
Sample size calculation
The sample size calculation was performed according to the primary outcome "Occurrence of reduced heart rate variability (RHRV)". This is to be estimated among subjects free from cardiac disease. We assumed that those make up about 50% in all age-sex strata. Assuming a true RHRV of 10%, a two-sided 95% confidence interval for this prevalence will have a one-sided length of 5.6% [4.4%; 15.6%] with 110 subjects free form cardiac disease. As such, a total of 220 subjects per 10-year-age-sex-stratum will be included, and a total sample size of 1760 subjects is needed for eight strata.
Discussion
In the ageing populations of industrialized nations, the increasing burden of chronic cardiovascular diseases already has an enormous impact on population health, the health care system and the economy. The need for a better understanding of how to achieve "healthy ageing", how to slow down the processes of cardiovascular disease generation and progression, and how to improve preventive and therapeutic strategies is obvious in societies with a steadily rising life expectancy. But a better understanding of the disease processes is also essential for the reversal of the deleterious effect of the political and economic disruptions in Eastern Europe since the late 1980s on the population health with sharply increasing mortality and decreasing life expectancy. There is increasing evidence for an important contribution of the dysbalance of the autonomic nervous system with predominance of the sympathetic system to disease processes. There are conflicting results regarding a possible mediating role of autonomic dysfunction on the pathway from socioeconomic and psychosocial risk factors to cardiovascular disease, but population-based data on HRV are scarce [39-42,79,80].
The CARLA study aims to create a database for a detailed analysis of the association of markers of autonomic dysfunction with other cardiovascular risk factors on the path to cardiovascular disease. The wide age range of the present study population, being representative of a general population, permits the analysis of age-dependent effects and of processes which become more important at older age. Due to the selection of standardized examination procedures and interview items applied in other regional studies, this study serves for comparative analyses which might increase the chance to discover pathways responsible for the CVD epidemic in some populations. The recording of 20-minute ECGs in a large sample of the elderly general population gives the unique opportunity to examine markers of autonomic dysfunction with a higher precision than the shorter ECGs recorded in most other population-based studies to date. The additional comparison with 24-hour ECGs in the same study population increases the diagnostic security of the short-term ECGs. The study of further parameters of ECG physiology could create a deeper understanding of pathophysiologic processes responsible for the development and progression of cardiovascular diseases. This is a prerequisite of targeted preventive and therapeutic strategies needed in populations with a growing percentage of elderly persons with naturally high disease rates or in populations with increased numbers of premature cardiovascular disease.
Conclusion
With the analysis of potential pathways from established and newer cardiovascular risk factors to cardiovascular disease, the CARLA study contributes important information needed for a successful change towards healthier ageing which can only be achieved by strengthening prevention as well as improving detection and supporting therapeutic steps in the management of cardiovascular diseases.
List of abbreviations
ARIC = Atherosclerosis Risk in Communities
ATC codes = Anatomic Therapeutic Classification
BIPS = Bremen Institute for Prevention Research and Social Medicine
CARLA = CARdiovascular disease, Living and Ageing in Halle
CES-D = Center of Epidemiological Studies Depression Scale
CHD = Coronary Heart Disease
CHF = Chronic Heart Failure
CVD = Cardiovascular Disease
ECG = Electrocardiogram
EPIC = European Prospective Investigation into Cancer and Nutrition
GKV = Gesetzliche Krankenversicherung (Compulsory Health Insurance)
GSF = National Research Center for Environment and Health, Neuherberg, Germany
HAPIEE = Health, Alcohol and Psychosocial Factors in Eastern Europe
HRV = Heart rate variability
IDOM = computer-based system of medication recording using the official GKV medication database
IL-1, IL-6 = interleukin-1, interleukin-6
LEADS = Interactive research-oriented ECG/VCG analysis system
KORA = Cooperative Health Research in the Augsburg Region
MEANS = Modular ECG Analysis System
MONICA = Monitoring Trends and Determinants in Cardiovascular Disease
NT-proBPN = n-terminal pro Brain Natriuretic Peptide
PZN = central pharmanumber
SF 12 = Short Form Health Survey Questionnaire
SHIP = Study of Health in Pomerania
SNP = Single Nucleotide Polymorphism
TNF-α = tumornecrosis factor α
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
KHG conceived of the study, designed major parts of the study, trained the study personnel, coordinates the study, participates in the statistical analyses and drafted the manuscript.
AK helps coordinate the study, participates in the statistical analyses and helped drafting the manuscript.
BS helped designing the interview, helps coordinating the study, participates in the statistical analyses and helped drafting the manuscript.
JAK helped drafting the manuscript and performs the Minnesota coding and pre-processing of ECGs for HRV analysis and contributed considerably to the MEANS software.
CAS helped designing the ECG recording procedure and drafting the manuscript, designed the HRV software and the LEADS software, and performs the HRV analyses.
OK helped drafting the manuscript, selecting the statistical procedures and participates in the statistical analyses.
JH helped designing the study and drafting the manuscript.
KW helped designing the study, coordinating the echocardiographic examinations, and drafting the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This study is funded by a grant from the Deutsche Forschungsgemeinschaft (DFG, German Research Foundation) as part of the Collaborative Research Center 598 "Heart failure in the elderly – cellular mechanisms and therapy" at the Medical Faculty of the Martin-Luther-University Halle-Wittenberg; by a grant of the Wilhelm-Roux Programme of the Martin-Luther-University Halle-Wittenberg; and by the Federal Employment Office. The test kits for the analysis of serum hs-CRP and NT-proBNP are provided to the central laboratory at the Leipzig University Clinics by Roche Diagnostics, Mannheim.
We would like to thank the study personnel Erika Kaiser, Ingrid Pfeil, Claudia Schönau, Daniela Koch, Barbara Piehler, Annett Malcherczyk, Kathrin Pieger, Heike Rauchhaus, Susann Koppsieker for their contribution to examining and interviewing the study subjects and for the data management. We also thank Alexander Jelzow, Rochus Witthaut, Barbara Panzner, Hendrik Schmidt, Frithjof Schlegel, Robert-Rainer Flieger, Matthias Rauchhaus, Konstantin M. Heinroth (Department of Medicine III, Martin-Luther-University Halle-Wittenberg) for their participation in the echocardiographic examination of the study subjects. We thank Annekatrin Bergmann and Ulrike Jendrezok for helping with the patient reports and Beate Jenderka for her invaluable secretarial and administrative support (Institute of Medical Epidemiology, Biostatistics and Informatics, Martin-Luther-University Halle).
We would like to thank Harald Loppnow for the laboratory analysis of cytokines (Department of Medicine III, Martin-Luther-University Halle-Wittenberg).
We would like to thank Joachim Thiery, Mathias Brügel, Matthias Orth, Daniel Teupser (ILM Leipzig) for their advice regarding the selection of laboratory parameters for analyses and for the analysis of those parameters.
We are indebted to Jan Lüdemann and Henry Völzke (SHIP Study Greifswald), Christa Meisinger, Marion Pietsch, Siegfried Perz, Rolf Holle, H. Nagl (KORA Study Augsburg), Kerstin Klipstein-Grobusch (EPIC Potsdam Study), Martin Bobak, Hynek Pikhart (HAPIEE Study, University College London), Jacqueline Witteman (Rotterdam Study), Achim Reineke (Bremer Institut für Präventionsforschung und Sozialmedizin) for their valuable support in selecting and implementing the recruitment, interview and examination methods and tools.
==== Refs
Association AH Heart Disease and Stroke. Statistics - 2005 Update 2005
Mosterd A Hoes AW Grobbee DE Epidemiology of heart failure: contours of an impending epidemic? Neth J Med 1998 53 235 244 9883000 10.1016/S0300-2977(98)00108-9
Mosterd A Hoes AW de Bruyne MC Deckers JW Linker DT Hofman A Grobbee DE Prevalence of heart failure and left ventricular dysfunction in the general population; The Rotterdam Study Eur Heart J 1999 20 447 455 10213348 10.1053/euhj.1998.1239
Raymond I Pedersen F Steensgaard-Hansen F Green A Busch-Sorensen M Tuxen C Appel J Jacobsen J Atar D Hildebrandt P Prevalence of impaired left ventricular systolic function and heart failure in a middle aged and elderly urban population segment of Copenhagen Heart 2003 89 1422 1429 14617553 10.1136/heart.89.12.1422
Roger VL Weston SA Redfield MM Hellermann-Homan JP Killian J Yawn BP Jacobsen SJ Trends in heart failure incidence and survival in a community-based population JAMA 2004 292 344 350 15265849 10.1001/jama.292.3.344
Haldeman GA Croft JB Giles WH Rashidee A Hospitalization of patients with heart failure: National Hospital Discharge Survey, 1985 to 1995 Am Heart J 1999 137 352 360 9924171 10.1053/hj.1999.v137.95495
O'Connell JB The economic burden of heart failure Clin Cardiol 2000 23 III6 10 10754775
O'Connor CM Friesinger GC Topol E Aging and the heart Textbook of cardiovascular medicine 1998 Philadelphia, Lippincott-Raven Publishers 817 840
Stewart S MacIntyre K Capewell S McMurray JJ Heart failure and the aging population: an increasing burden in the 21st century? Heart 2003 89 49 53 12482791 10.1136/heart.89.1.49
Bobak M Marmot M East-West mortality divide and its potential explanations: proposed research agenda BMJ 1996 312 421 425 8601115
Walberg P McKee M Shkolnikov V Chenet L Leon DA Economic change, crime, and mortality crisis in Russia: regional analysis BMJ 1998 317 312 318 9685275
WHO Ecological analysis of the association between mortality and major risk factors of cardiovascular disease. The World Health Organization MONICA Project Int J Epidemiol 1994 23 505 516 7960374
Willich S Löwel H Mey W Trautner C Regional Variations in Mortality of Cardiovascular Diseases in Germany. Deutsches Ärzteblatt 1999 96 A483 A488
Hemingway H Marmot M al. YS Psychosocial factors in the primary and secondary prevention of coronary heart disease: a systematic review. Evidence based cardiology 1998 London, BMJ Books 269 286
Kristenson M Orth-Gomer K Kuchinskiene Z Hertzman C and al Different patterns of psychosocial strain; a possible explanation for the differences in ischemic heart disease mortality between Sweden and Lithuania? East-West Life Expectancy Gap in Europe 1996 Kluwer Academic Publishers 187 194
Peter R Alfredsson L Hammar N Siegrist J Theorell T Westerholm P High effort, low reward, and cardiovascular risk factors in employed Swedish men and women: baseline results from the WOLF Study J Epidemiol Community Health 1998 52 540 547 10320854
Ferketich AK Schwartzbaum JA Frid DJ Moeschberger ML Depression as an antecedent to heart disease among women and men in the NHANES I study. National Health and Nutrition Examination Survey Arch Intern Med 2000 160 1261 1268 10809028 10.1001/archinte.160.9.1261
Stansfeld SA Fuhrer R Shipley MJ Marmot MG Psychological distress as a risk factor for coronary heart disease in the Whitehall II Study Int J Epidemiol 2002 31 248 255 11914328 10.1093/ije/31.1.248
Rosengren A Hawken S Ounpuu S Sliwa K Zubaid M Almahmeed WA Blackett KN Sitthi-amorn C Sato H Yusuf S Association of psychosocial risk factors with risk of acute myocardial infarction in 11119 cases and 13648 controls from 52 countries (the INTERHEART study): case-control study Lancet 2004 364 953 962 15364186 10.1016/S0140-6736(04)17019-0
Marmot MG Smith GD Stansfeld S Patel C North F Head J White I Brunner E Feeney A Health inequalities among British civil servants: the Whitehall II study Lancet 1991 337 1387 1393 1674771 10.1016/0140-6736(91)93068-K
Marmot MG Bosma H Hemingway H Brunner E Stansfeld S Contribution of job control and other risk factors to social variations in coronary heart disease incidence Lancet 1997 350 235 239 9242799 10.1016/S0140-6736(97)04244-X
Labarthe DR Adverse Psychosocial Patterns Epidemiology and Prevention of Cardiovascular Diseases A Global Challenge 1998 Gaithersburg, MD, Aspen Publishers, Inc. 393 409
Labarthe DR Social Conditions Epidemiology and Prevention of Cardiovascular Diseases A Global Challenge 1998 Gaithersburg, MD, Aspen Publishers, Inc. 423 447
Liao D Barnes RW Chambless LE Simpson RJJ Sorlie P Heiss G Age, race, and sex differences in autonomic cardiac function measured by spectral analysis of heart rate variability--the ARIC study. Atherosclerosis Risk in Communities Am J Cardiol 1995 76 906 912 7484830 10.1016/S0002-9149(99)80260-4
Agelink MW Malessa R Baumann B Majewski T Akila F Zeit T Ziegler D Standardized tests of heart rate variability: normal ranges obtained from 309 healthy humans, and effects of age, gender, and heart rate Clin Auton Res 2001 11 99 108 11570610
Tsuji H Venditti FJJ Manders ES Evans JC Larson MG Feldman CL Levy D Reduced heart rate variability and mortality risk in an elderly cohort. The Framingham Heart Study Circulation 1994 90 878 883 8044959
Liao D Cai J Rosamond WD Barnes RW Hutchinson RG Whitsel EA Rautaharju P Heiss G Cardiac autonomic function and incident coronary heart disease: a population-based case-cohort study. The ARIC Study. Atherosclerosis Risk in Communities Study Am J Epidemiol 1997 145 696 706 9125996
Huikuri HV Makikallio TH Airaksinen KE Seppanen T Puukka P Raiha IJ Sourander LB Power-law relationship of heart rate variability as a predictor of mortality in the elderly Circulation 1998 97 2031 2036 9610533
de Bruyne MC Kors JA Hoes AW Klootwijk P Dekker JM Hofman A van Bemmel JH Grobbee DE Both decreased and increased heart rate variability on the standard 10-second electrocardiogram predict cardiac mortality in the elderly: the Rotterdam Study Am J Epidemiol 1999 150 1282 1288 10604770
Dekker JM Crow RS Folsom AR Hannan PJ Liao D Swenne CA Schouten EG Low heart rate variability in a 2-minute rhythm strip predicts risk of coronary heart disease and mortality from several causes: the ARIC Study. Atherosclerosis Risk In Communities Circulation 2000 102 1239 1244 10982537
Bigger JT Fleiss JL Rolnitzky LM Steinman RC The ability of several short-term measures of RR variability to predict mortality after myocardial infarction Circulation 1993 88 927 934 8353919
Tsuji H Larson MG Venditti FJJ Manders ES Evans JC Feldman CL Levy D Impact of reduced heart rate variability on risk for cardiac events. The Framingham Heart Study Circulation 1996 94 2850 2855 8941112
Nolan J Batin PD Andrews R Lindsay SJ Brooksby P Mullen M Baig W Flapan AD Cowley A Prescott RJ Neilson JM Fox KA Prospective study of heart rate variability and mortality in chronic heart failure: results of the United Kingdom heart failure evaluation and assessment of risk trial (UK-heart) Circulation 1998 98 1510 1516 9769304
Janszky I Ericson M Mittleman MA Wamala S Al Khalili F Schenck-Gustafsson K Orth-Gomer K Heart rate variability in long-term risk assessment in middle-aged women with coronary heart disease: The Stockholm Female Coronary Risk Study J Intern Med 2004 255 13 21 14687234 10.1046/j.0954-6820.2003.01250.x
Horsten M Ericson M Perski A Wamala SP Schenck-Gustafsson K Orth-Gomer K Psychosocial factors and heart rate variability in healthy women Psychosom Med 1999 61 49 57 10024067
Carney RM Blumenthal JA Stein PK Watkins L Catellier D Berkman LF Czajkowski SM O'Connor C Stone PH Freedland KE Depression, heart rate variability, and acute myocardial infarction Circulation 2001 104 2024 2028 11673340
Sloan RP Bagiella E Shapiro PA Kuhl JP Chernikhova D Berg J Myers MM Hostility, gender, and cardiac autonomic control Psychosom Med 2001 63 434 440 11382270
Virtanen R Jula A Salminen JK Voipio-Pulkki LM Helenius H Kuusela T Airaksinen J Anxiety and hostility are associated with reduced baroreflex sensitivity and increased beat-to-beat blood pressure variability Psychosom Med 2003 65 751 756 14508016 10.1097/01.PSY.0000088760.65046.CF
Carney RM Blumenthal JA Freedland KE Stein PK Howells WB Berkman LF Watkins LL Czajkowski SM Hayano J Domitrovich PP Jaffe AS Low heart rate variability and the effect of depression on post-myocardial infarction mortality Arch Intern Med 2005 165 1486 1491 16009863 10.1001/archinte.165.13.1486
Hemingway H Shipley M Brunner E Britton A Malik M Marmot M Does autonomic function link social position to coronary risk? The Whitehall II study Circulation 2005 111 3071 3077 15939818 10.1161/CIRCULATIONAHA.104.497347
Janszky I Ericson M Blom M Georgiades A Magnusson JO Alinagizadeh H Ahnve S Wine drinking is associated with increased heart rate variability in women with coronary heart disease Heart 2005 91 314 318 15710709 10.1136/hrt.2004.035105
Steptoe A Feldman PJ Kunz S Owen N Willemsen G Marmot M Stress responsivity and socioeconomic status: a mechanism for increased cardiovascular disease risk? Eur Heart J 2002 23 1757 1763 12419295 10.1053/euhj.2001.3233
Vasan RS Sullivan LM Roubenoff R Dinarello CA Harris T Benjamin EJ Sawyer DB Levy D Wilson PW D'Agostino RB Inflammatory markers and risk of heart failure in elderly subjects without prior myocardial infarction: the Framingham Heart Study Circulation 2003 107 1486 1491 12654604 10.1161/01.CIR.0000057810.48709.F6
Ingelsson E Arnlov J Sundstrom J Lind L Inflammation, as measured by the erythrocyte sedimentation rate, is an independent predictor for the development of heart failure J Am Coll Cardiol 2005 45 1802 1806 15936609 10.1016/j.jacc.2005.02.066
Sajadieh A Nielsen OW Rasmussen V Hein HO Abedini S Hansen JF Increased heart rate and reduced heart-rate variability are associated with subclinical inflammation in middle-aged and elderly subjects with no apparent heart disease Eur Heart J 2004 25 363 370 15033247 10.1016/j.ehj.2003.12.003
Heart rate variability. Standards of measurement, physiological interpretation, and clinical use. Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology Eur Heart J 1996 17 354 381 8737210
Dekker JM De Vries EL Lengton RR Maan AC Schouten EG Swenne CA Maan A Reproducibility and comparability of short- and long-term heart rate variability measures in healthy young men. Ann Noninvasive Electrocardiol 1996 1 287 292
Hense HW Filipiak B Döring A Stieber J Liese A Keil U Ten-year trends of cardiovascular risk factors in the MONICA Augsburg Region in Southern Germany. Results from the 1984/85, 1989/90 and 1994/1995 surveys. CVD Prevention 1998 1 318 327
John U Greiner B Hensel E Ludemann J Piek M Sauer S Adam C Born G Alte D Greiser E Haertel U Hense HW Haerting J Willich S Kessler C Study of Health In Pomerania (SHIP): a health examination survey in an east German region: objectives and design Soz Praventivmed 2001 46 186 194 11565448 10.1007/BF01324255
Kroke A Klipstein-Grobusch K Voss S Moseneder J Thielecke F Noack R Boeing H Validation of a self-administered food-frequency questionnaire administered in the European Prospective Investigation into Cancer and Nutrition (EPIC) Study: comparison of energy, protein, and macronutrient intakes estimated with the doubly labeled water, urinary nitrogen, and repeated 24-h dietary recall methods Am J Clin Nutr 1999 70 439 447 10500011
Hofman A Grobbee DE de Jong PT van den Ouweland FA Determinants of disease and disability in the elderly: the Rotterdam Elderly Study Eur J Epidemiol 1991 7 403 422 1833235 10.1007/BF00145007
Bobak M Room R Pikhart H Kubinova R Malyutina S Pajak A Kurilovitch S Topor R Nikitin Y Marmot M Contribution of drinking patterns to differences in rates of alcohol related problems between three urban populations J Epidemiol Community Health 2004 58 238 242 14966239 10.1136/jech.2003.011825
Giesecke B Nagl H GmbH GSFFUG DAIMON -Rechnergestützte Fragebogenentwicklung und Interviewführung. Dokumentation. GSF-Bericht 03/01. 2001 GSF-Bericht 03/01 Neuherberg, GSF-Forschungszentrum 1 91
Rose G McCartney P Reid DD Self-administration of a questionnaire on chest pain and intermittent claudication Br J Prev Soc Med 1977 31 42 48 856370
Döring A Fischer B Holle R Hoppe S Immervoll T Janssen C John J Merkl J Nagl H Mühlberger N Papke K Perz S Pietsch M Rathmann W Schäfer T Schwertner B Stieber J Zahlmann G KORA-Survey 2000. Manual of Operation. Untersucher-Handbuch. 2000 GSF Neuherberg
Fricke U Günther J Zawinell A Anatomisch-therapeutisch-chemische Klassifikation mit Tagesdosen für den deutschen Arzneimittelmarkt. Methodik der ATC-Klassifikation und DDD-festlegung. ATC-Index mit DDD-Angaben. GKV-Arzneimittelindex 2005 4. Bonn, Wissenschaftliches Institut der Ortskrankenkassen (WIdO) 15590332
Siegrist J Adverse health effects of high-effort/low-reward conditions J Occup Health Psychol 1996 1 27 41 9547031 10.1037/1076-8998.1.1.27
Siegrist J Starke D Chandola T Godin I Marmot M Niedhammer I Peter R The measurement of effort-reward imbalance at work: European comparisons Soc Sci Med 2004 58 1483 1499 14759692 10.1016/S0277-9536(03)00351-4
O'Brien E Mee F Atkins N Thomas M Evaluation of three devices for self-measurement of blood pressure according to the revised British Hypertension Society Protocol: the Omron HEM-705CP, Philips HP5332, and Nissei DS-175 Blood Press Monit 1996 1 55 61 10226203
Mundt KA Chambless LE Burnham CB Heiss G Measuring ankle systolic blood pressure: validation of the Dinamap 1846 SX Angiology 1992 43 555 566 1626733
National Heart Lung and Blood Institute of the National Institutes of Health ARIC Manuals of Operation. Manual 5 Electrocardiography 1987 5 Chapel Hill, NC, ARIC Coordinating Center, School of Public Health, University of North Carolina at Chapel Hill
WHO MONICA Project. The MONICA Manual. 1990 Geneva, WHO Cardiovascular Disease Unit
Piek M Lüdeman J Völzke H Hummel A Hense HW Operationshandbuch Echokardiographie. Study of Health in Pomerania (SHIP-I). 1. Follow-up-Untersuchung 2002 SHIP-I Organisationszentrum, Ernst-Moritz-Arndt-Universität Greifswald, Walter-Rathenau-Str. 48, 17489 Geifswald
Thomas L Labor und Diagnose 2005 6. TH-Books
Kimberly MM Leary ET Cole TG Waymack PP Selection, validation, standardization, and performance of a designated comparison method for HDL-cholesterol for use in the cholesterol reference method laboratory network Clin Chem 1999 45 1803 1812 10508128
Eda S Kaufmann J Roos W Pohl S Development of a new microparticle-enhanced turbidimetric assay for C-reactive protein with superior features in analytical sensitivity and dynamic range J Clin Lab Anal 1998 12 137 144 9591699 10.1002/(SICI)1098-2825(1998)12:3<137::AID-JCLA2>3.0.CO;2-6
Richards AM Nicholls MG Yandle TG Frampton C Espiner EA Turner JG Buttimore RC Lainchbury JG Elliott JM Ikram H Crozier IG Smyth DW Plasma N-terminal pro-brain natriuretic peptide and adrenomedullin: new neurohormonal predictors of left ventricular function and prognosis after myocardial infarction Circulation 1998 97 1921 1929 9609085
Struthers AD How to use natriuretic peptide levels for diagnosis and prognosis Eur Heart J 1999 20 1374 1375 10487796 10.1053/euhj.1999.1771
van Bemmel JH Kors JA van Herpen G Methodology of the modular ECG analysis system MEANS Methods Inf Med 1990 29 346 353 2233382
Prineas RJ Crow. RS Blackburn H The Minnesota Code Manual of Electrocardiographic Findings Standard Procedures for Measurement and Classification 1982 Boston, John Wright PSB
Kors JA van Herpen G Wu J Zhang Z Prineas RJ van Bemmel JH Validation of a new computer program for Minnesota coding J Electrocardiol 1996 29 Suppl 83 88 9238383 10.1016/S0022-0736(96)80025-2
de Bruyne MC Kors JA Hoes AW Kruijssen DA Deckers JW Grosfeld M van Herpen G Grobbee DE van Bemmel JH Diagnostic interpretation of electrocardiograms in population-based research: computer program research physicians, or cardiologists? J Clin Epidemiol 1997 50 947 952 9291880 10.1016/S0895-4356(97)00100-5
Kors JA Crow RS Hannan PJ Rautaharju PM Folsom AR Comparison of computer-assigned Minnesota Codes with the visual standard method for new coronary heart disease events Am J Epidemiol 2000 151 790 797 10965976
Bootsma M Swenne CA Van Bolhuis HH Chang PC Cats VM Bruschke AV Heart rate and heart rate variability as indexes of sympathovagal balance Am J Physiol 1994 266 H1565 H1571 8184935
Pluim BM Swenne CA Zwinderman AH Maan AC van der LA Doornbos J Van der Wall EE Correlation of heart rate variability with cardiac functional and metabolic variables in cyclists with training induced left ventricular hypertrophy Heart 1999 81 612 617 10336920
Priori SG Mortara DW Napolitano C Diehl L Paganini V Cantu F Cantu G Schwartz PJ Evaluation of the spatial aspects of T-wave complexity in the long-QT syndrome Circulation 1997 96 3006 3012 9386169
Mirvis DM Goldberger AL Zipes DP, Libby P, Bonow RO and Braunwald E Electrocardiography Braunwald's Heart Disease 2005 Ch. 9 7th Edition Philadelphia, Elsevier Saunders 107 151
Draisma HHM Swenne CA Van de Vooren H Maan AC Hooft van huysduynen B Van der Wall EE Schalij MJ LEADS, an interactive research-oriented ECG/VCG analysis system Computers in Cardiology 2005 32 accepted for publication, 4 pp.
Gallo LC Bogart LM Vranceanu AM Walt LC Job characteristics, occupational status, and ambulatory cardiovascular activity in women Ann Behav Med 2004 28 62 73 15249260 10.1207/s15324796abm2801_8
Riese H Van Doornen LJ Houtman IL De Geus EJ Job strain in relation to ambulatory blood pressure, heart rate, and heart rate variability among female nurses Scand J Work Environ Health 2004 30 477 485 15635758
Jöckel KH Babitsch B Bellach BM Bloomfield K Hoffmeyer-Zlotnik J Winkler J Wolf C Ahrens W, Bellach BM and Jöckel KH Messung und Quantifizierung soziodemographischer Merkmale in epidemiologischen Studien Messung soziodemographischer Merkmale in der Epidemiologie 1998 München, RKI-Schriften 1/98; MMV Medizin Verlag GmbH 7 38
Adam C Alte D Born G Eichenauer-Rettig U John U Lüdemann J Paritschke H Piek M Sauer S Wussow T Härtel U Operationshandbuch Computergestütztes Interview. Study of Health in Pomerania (SHIP-I). 1. Follow-up-Untersuchung 2001 Greifswald, SHIP-I Organisationszentrum, Ernst-Moritz-Arndt-Universität Greifswald, Walter-Rathenau-Str. 48, 17489 Geifswald
Keil U Cairns V Döring A Härtel U Jorcik J Perz S Stieber J MONICA-Augsburg Manual of Operations - Survey. GSF-Bericht 20/85 1985 20/85 München, Gesellschaft für Strahlen- und Umweltforschung (GSF)
The World Health Organization MONICA Project (monitoring trends and determinants in cardiovascular disease): a major international collaboration. WHO MONICA Project Principal Investigators J Clin Epidemiol 1988 41 105 114 3335877 10.1016/0895-4356(88)90084-4
Bothig S WHO MONICA Project: objectives and design Int J Epidemiol 1989 18 S29 S37 2807705
Winkler G Döring A Kurzmethoden zur Charakterisierung des Ernährungsmusters: Einsatz und Auswertung eines Food-Frequency-Fragebogens Ernährungs-Umschau 1995 42 289 291
Baecke JA Burema J Frijters JE A short questionnaire for the measurement of habitual physical activity in epidemiological studies Am J Clin Nutr 1982 36 936 942 7137077
Berkman LF Syme SL Social networks, host resistance, and mortality: a nine-year follow-up study of Alameda County residents Am J Epidemiol 1979 109 186 204 425958
Seeman TE Berkman LF Structural characteristics of social networks and their relationship with social support in the elderly: who provides support Soc Sci Med 1988 26 737 749 3358145 10.1016/0277-9536(88)90065-2
Putnam RD Leonardi R Nanetti RY Making democracy work: Civic traditions in modern Italy 1993 Princeton, Princeton University Press
Rose R New Russia Barometer VI: After the presidential election. Studies in Public Policy, No. 272 1996 Glasgow, Center for the Study of Public Policy, University of Strathclyde
Rose R How free from fear are citizen in transition societies? 2001
Bullinger M Kirchberger I SF-36 Fragebogen zum Gesundheitszustand (Handanweisung) 1998 Göttingen, Bern, Toronto, Seattle, Hogrefe Verlag für Psychologie
Hautzinger M Die CES-D Skala. Ein Depressionsmessinstrument für Untersuchungen in der Allgemeinbevölkerung. Diagnostika 1988 34 167 173
Hautzinger M Bailer M Allgemeine Depressionsskala (ADS) 1996 Göttingen, Belt
Kohlmann TH Gerbershagen HU CES-D, Deutsche Version 2005
|
16283930
|
PMC1299322
|
CC BY
|
2021-01-04 16:30:07
|
no
|
BMC Cardiovasc Disord. 2005 Nov 11; 5:33
|
utf-8
|
BMC Cardiovasc Disord
| 2,005 |
10.1186/1471-2261-5-33
|
oa_comm
|
==== Front
BMC Evol BiolBMC Evolutionary Biology1471-2148BioMed Central London 1471-2148-5-681630067910.1186/1471-2148-5-68Research ArticlePhylogenetic analyses suggest reverse splicing spread of group I introns in fungal ribosomal DNA Bhattacharya Debashish [email protected] Valérie [email protected] Dawn M [email protected] François [email protected] Department of Biological Sciences and Roy J. Carver Center for Comparative Genomics, University of Iowa, 446 Biology Building, Iowa City, IA 52242-1324, USA2 Department of Biology, Duke University, Durham, NC 27708-0338, USA3 Department of Biological Sciences, University of Calgary, Calgary, Alberta T2N 1N4, Canada2005 21 11 2005 5 68 68 24 7 2005 21 11 2005 Copyright © 2005 Bhattacharya et al; licensee BioMed Central Ltd.2005Bhattacharya et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Group I introns have spread into over 90 different sites in nuclear ribosomal DNA (rDNA) with greater than 1700 introns reported in these genes. These ribozymes generally spread through endonuclease-mediated intron homing. Another putative pathway is reverse splicing whereby a free group I intron inserts into a homologous or heterologous RNA through complementary base-pairing between the intron and exon RNA. Reverse-transcription of the RNA followed by general recombination results in intron spread. Here we used phylogenetics to test for reverse splicing spread in a taxonomically broadly sampled data set of fungal group I introns including 9 putatively ancient group I introns in the rDNA of the yeast-like symbiont Symbiotaphrina buchneri.
Results
Our analyses reveal a complex evolutionary history of the fungal introns with many cases of vertical inheritance (putatively for the 9 introns in S. buchneri) and intron lateral transfer. There are several examples in which introns, many of which are still present in S. buchneri, may have spread through reverse splicing into heterologous rDNA sites. If the S. buchneri introns are ancient as we postulate, then group I intron loss was widespread in fungal rDNA evolution.
Conclusion
On the basis of these results, we suggest that the extensive distribution of fungal group I introns is at least partially explained by the reverse splicing movement of existing introns into ectopic rDNA sites.
==== Body
Background
Group I introns are autocatalytic RNAs that are widespread in organellar and nuclear genomes of eukaryotes, in eubacteria, and in phages and viruses [reviewed in [1-4]]. How these elements "move" within and between genes and between natural populations and species is poorly understood [4,5]. Two mechanisms are invoked to explain group I intron spread. The first is homing and is initiated by an intron-encoded endonuclease (homing endonuclease gene [HEG]) that recognizes and cleaves an intron-less allele at or near the intron insertion site [reviewed in [6]]. Following endonuclease cleavage at a specific 15 – 20 nt target sequence, the intron-containing allele is used as the template in a double-strand break repair pathway resulting in insertion of the intron and co-conversion of flanking exon sequences [7,8]. HEGs appear to be recurrently gained, degenerate, and lost in a cyclical manner and the intron-HEG combination is eventually lost from a population after all individuals are fixed for these elements [9]. A recent analysis in our lab of HEGs in nuclear rDNA group I introns showed that these coding regions are mobile elements (as has been shown for many organellar group I introns [reviewed in [6]]) that move either to introns in homologous rDNA sites or to introns in neighboring sites in often evolutionarily distantly related species [4,10,11]. The invading HEGs can then mobilize their new intron partners and achieve rapid spread within populations. However, our in-depth phylogenetic analyses of the HEGs failed to show the involvement of these endonucleases in the movement of group I introns into distant rDNA sites [10]. This type of long-distance (e.g., >50 nt) movement has, however, been suggested by phylogenetic studies that show group I introns from sites such as SSU rDNA S287 and S1199 (numbering based on the Escherichia coli gene) to be closely related [12].
Long-distance intron movement can in principle be achieved by reverse splicing that facilitates intron mobility through an RNA intermediate [13-16]. In reverse splicing, group I introns recognize their target sequence through complementary base pairing with a short (4–6 nt) internal guide sequence (see Fig. 1) followed by integration into the transcript [e.g., [14]] and then putatively reverse-transcription, and general recombination to achieve spread. The importance of this pathway in group I intron movement in nature, however, remains to be established because reverse splicing-mediated intron movement has not been demonstrated in genetic crosses. Furthermore, whereas homing is highly efficient in spreading introns in populations, it is likely that reverse splicing with its reliance on chance integration followed by two additional steps (i.e., reverse-transcription, recombination) would be less efficient in promoting intron movement. An additional constraint is that rDNA exists as a multi-copy gene family necessitating that alleles containing transferred introns must rise to high frequency (presumably through concerted evolution or less parsimoniously, repeated reverse splicing events) in individuals and in populations to ensure survival and ultimately, fixation. If group I introns are weakly deleterious, then fixation may occur only in species with small population sizes [17]. These considerations suggest that rare reverse splicing events may be most successfully recognized in the context of broadly sampled host and intron phylogenies in which many potential candidates for reverse splicing movement are studied. The large collection of fungal group I introns that has recently accumulated provides an ideal opportunity to test comprehensively the contribution of reverse splicing to the extant intron distribution.
Figure 1 Group I intron splicing. A) The typical secondary structure of a group I intron which consists of about ten paired elements (P1–P10). The intron internal guide sequence (IGS) is shown that recognizes the 5' exon sequence through a 4–6 nt base pairing, thereby initiating the two-step splicing mechanism (shown in panel B) for group I intron removal from pre-RNA. B) The forward and reverse splicing of a group I intron is shown in this figure. The arrows that lead from the free intron to the pre-RNA indicate the reverse splicing reaction. Both forward and reverse splicing reactions require the IGS interaction in domain P1.
To assess whether phylogenetic evidence exists that is consistent with a reverse splicing mechanism of group I intron spread, we analyzed a large group I intron data set (189 sequences) from the Pezizomycotina. These fungi are particularly intron-rich with some taxa containing up to 10 ribozymes in the rDNA (e.g., Physconia perisidiosa). In addition, we included 10 rDNA group I introns from the yeast-like fungal symbiont of anobiid beetles, Symbiotaphrina spp. [18]. We addressed two questions with this study: 1) Is there phylogenetic evidence for the reverse splicing-mediated spread of group I introns in the Pezizomycotina fungi? 2) Are the multiple group I introns in Symbiotaphrina spp. that are shared with other fungi the result of independent lateral transfers into the rDNA genes of these taxa, or have some or all of these introns been vertically inherited in the fungi? For the latter question we were particularly interested in determining whether Symbiotaphrina, which lives in the gut of beetles and could therefore theoretically come in contact with many different cells, be a potential vector for group I intron spread. Although many studies have provided evidence for the movement of group I and group II introns across broad evolutionary lines [e.g., [5,12,19-22]], there is little known about the vectors that facilitate their spread. Our analyses demonstrate a complex evolutionary history of the fungal introns with many cases of vertical inheritance (putatively for the 9 introns in S. buchneri) and intron lateral transfer. In addition, there are several examples in which introns, many of which are still present in S. buchneri, may have spread through reverse splicing into heterologous rDNA sites.
Results and discussion
General patterns of fungal group I intron inheritance
We took advantage of the most group I intron-rich of all eukaryotes, the Pezizomycotina fungi (in particular, the lichen-forming fungi [12,23]), to address the movement and long-term evolution of these mobile elements. The nuclear group I introns are found exclusively in rDNA genes with some taxa containing 7 (Gymnoderma coccocarpum), 8 (Diplotomma epipolium), 9 (Symbiotaphrina buchneri), or 10 insertions (Physconia perisidiosa [12,23]) in their SSU and LSU rDNA. To facilitate the analysis of fungal group I introns, we first inferred a "host" tree of the Pezizomycotina based on the analysis of SSU and LSU rDNA + RPB2 data (Fig. 2). The rDNA sequences for the four S. kochii strains were identical (each encoded a single group I intron at position L1921). The rDNA coding regions of S. buchneri JCM9740 were interrupted by 5 introns in the SSU rDNA coding region (S114, S287, S1052, S1210, S1506) and by 4 introns in the LSU rDNA (L1094, L1921, L2066, L2449). In the host tree, S. buchneri and S. kochii form a clade with Bayesian but not bootstrap support, providing weak evidence for their monophyly (see grey box in Fig. 2). This result was previously found in a more limited analysis of partial small subunit rDNA sequences [24]. In our tree, the Symbiotaphrina species diverge before the split of most of the major lineages of lichen fungi (i.e., Lecanoromycetes + Eurotiomycetes), but again without bootstrap support. If the early divergence of Symbiotaphrina is correct, then the fungal host tree suggests that under a model of vertical inheritance, in intron trees, the nine S. buchneri group I introns should each form independent monophyletic groups containing other fungal introns at the respective insertion sites. The S. buchneri sequences should not cluster on the basis of their common occurrence in this species and ideally, they should branch after the Sordariomycetes but before the Lecanoromycetes + Eurotiomycetes group (see Fig. 2).
Figure 2 Host tree of fungi. Bayesian and NJ analyses of fungi showing the position of the genus Symbiotaphrina within the Ascomycota. This tree is inferred from a combined data set of nuclear SSU rDNA, LSU rDNA, and RPB2 from 84 species of the Ascomycota with one basidiomycete species used as the outgroup. The phylogram represents the majority rule consensus tree of 40,000 post-burnin trees sampled by the Bayesian search algorithm. The lengths for each branch were averaged over all trees having this branch (sumt option in MrBayes v30b4). Numbers above internodes are posterior probabilities (when ≥95%). Values below the internodes are NJ bootstrap proportions. If both the Bayesian posterior probabilities are ≥95% and the NJ bootstrap support are ≥70%, the internal branch is shown as a thicker line. The grey box delimits the genus Symbiotaphrina (bluish green text). Supra-generic taxon names follow [49]. The major intron-containing fungal groups are shown in different colors (Acarosporomycetidae in blue, Ostropomycetidae in vermillion, Lecanoromycetidae in orange, and the Sordariomycetes in reddish purple).
Given these expectations, we first analyzed the 189-sequence intron data set that was used as input for the JC-NJ and Bayesian inference of phylogeny (results not shown). This tree was consistent with a model of vertical inheritance with the fungal introns assorted primarily on the basis of their sites of rDNA insertion and not intermixed, which would result if they had been frequently laterally transferred to ectopic sites ([3,10,12,22,25] for exceptions, see below). A JC-NJ analysis using a subset of 116 sequences provided the same results (Fig. 3) but with higher bootstrap support due to the reduced number of taxa in the data set. In particular, the nine S. buchneri group I introns (green filled triangles in Fig. 3) are distributed in the tree on the basis of their site of rDNA genic insertion, often in clades with bootstrap and/or Bayesian support. The same holds for the P. perisidiosa group I introns (yellow filled circles) in Fig. 3. These results argue for a separate evolutionary history for the different introns although within each intron lineage there could be a combination of vertical inheritance and lateral transfer between fungi [e.g., [10]]. This would result if some of the introns were targeted (i.e., fixed) at the homologous rDNA site in other species resulting in their monophyly; i.e., supporting intron vertical inheritance even though the ribozymes had moved between species. This phenomenon is best addressed with targeted analyses of specific intron lineages and the fungi that contain these introns [e.g., [23,25]].
Figure 3 Phylogeny of group I introns. Phylogeny of fungal group I introns implicated in reverse splicing movement. This is a JC-NJ tree that was inferred for 116 fungal introns. The results of a JC-NJ bootstrap analysis are shown above the branches, whereas the results of an unweighted maximum parsimony bootstrap analysis are shown below the branches. The thick branches represent ≥95% Bayesian posterior probability. Branch lengths are proportional to the number of substitutions per site (see legend). The Symbiotaphrina spp. group I introns are marked with the filled triangles and the P. perisidiosa introns are marked with the filled circles. The intron insertion sites in small (S) and large (L) subunit rDNA are shown. The filled squares at the nodes denote the putative cases of intron movement. The colors for the different introns reflect the taxonomic position of the host cell containing the ribozymes (consistent with the scheme shown in Fig 2).
In general the introns in S. buchneri are either limited to the Lecanoromycetes (S114, S1052, S1210) or are more closely related to homologs in this group than in the Sordariomycetes, as would be expected under a model of vertical inheritance (e.g., L1921, L2449). The single L1921 group I intron in S. kochii appears however to have an origin through lateral transfer from a sordariomycetes source (albeit without bootstrap support, Fig. 3). We tested the hypothesis of independent origins of the S. buchneri group I introns by forcing the monophyly of these sequences from different sites in the tree inferred from the 51-sequence data set (see Fig. 4). The Shimodaira-Hasegawa statistical test significantly rejected trees, for example, in which the S. buchneri S1210 and S114 (S1210 branch moved to S114, or S114 branch moved to S1210, P < 0.01), the S287 and L2449 (S287 to L2449, or L2449 to S287, P < 0.001), or the S114 and S1506 (S114 to S1506, or S1506 to S114, P < 0.001) introns were united in one clade. Forcing the monophyly of any of the S. buchneri IC and IE introns resulted in the greatest differences in log likelihood arguing for a long evolutionary separation of these intron subgroups. The only tree rearrangement that did not result in significant SH-test scores was for the union of the S. buchneri S1052 and S1506 group I introns (S1052 to S1506, P = 0.153; S1506 to S1052, P = 0.195). This suggests a potential common evolutionary origin of these introns.
Figure 4 Testing putative cases of group I intron reverse splicing. Phylogeny of fungal introns and analysis of rDNA flanking exons to assess putative cases of group I intron reverse splicing. A) Distance matrix phylogenetic tree of a reduced data set of 51 fungal introns. The results of a distance bootstrap analysis are shown above the branches on the left of the slash marks, whereas the results of a maximum likelihood bootstrap analysis are shown on the right of the slash marks. The values shown below the branches result from an unweighted maximum parsimony bootstrap analysis. The thick branches represent ≥95% Bayesian posterior probability. Branch lengths are proportional to the number of substitutions per site (see legend). The S. buchneri group I introns are marked with filled triangles and the P. perisidiosa introns are marked with the filled circles. The putative cases of intron movement are denoted with the filled squares on the internal nodes. The rDNA intron insertion site is shown for each ribozyme. B) Majority-rule consensus tree inferred from a Bayesian analysis of 34 fungal group I introns. Only the core regions of the ribozymes were used in this analysis. The colors for the different introns in panels A and B reflect the taxonomic position of the host cell containing the ribozymes (consistent with the scheme shown in Fig 2). C) Alignment of exon sequences (all from S. buchneri) at heterologous group I intron sites. Regions required for the IGS interaction that are implicated in reverse splicing intron movement are shown in the boxed areas.
In summary, our data are consistent with the idea that the S. buchneri group I introns have independent origins and that they been vertically inherited in many fungi. Our data argue most strongly for the view that the S. buchneri introns have not recently spread in rDNA through ectopic transposition within this taxon. Consistent with these ideas, detailed analysis of the Sordariomycetes support the hypothesis of long-term group I intron vertical evolution [25]. We cannot, however, unambiguously ascertain the extent of intron lateral transfer between all Pezizomycotina because of the uneven and sporadic distribution of the data (i.e., the Sordariomycetes and Lecanoromycetes contain the majority of fungal group I introns. If the S. buchneri introns are ancient as we postulate, then group I intron loss was widespread in fungal rDNA evolution because these sequences are not present in many fungi (e.g., completely absent, to date, in the Dothideomycetidae and rare in the Eurotiomycetidae [Reeb et al. unpublished data]). Within the Lecanoromycetidae, detailed analysis of this group using SSU rDNA comparisons shows that entire derived lineages (e.g., Bacidiaceae, Peltigeraceae, Rhizocarpaceae) are intron-free even though their sisters often contain multiple different group I introns (e.g., Acarosporaceae [Reeb et al., unpublished data], Cladoniaceae, Physciaceae [12,23]).
Evidence for group I intron spread
Although group I introns form monophyletic groups in Figure 3, there are also cases of a close relationship (bootstrap and/or Bayesian support) between intron clades at ectopic rDNA sites. Inspection of Figures 3 and 4 suggests seven potential transposition events (marked with filled boxes). These same cases were also present in the 189-sequence data set (results not shown). The intron movements involve four cases that were previously found in more limited phylogenetic analyses of clades S114 – S303, S1046 – S1052, S287 – S1199 (without bootstrap or Bayesian support in Fig. 3, but see Fig. 4A), S1506 – S1516 [12] and three putative new cases found in this study (S497 – S1210, S934 – L1025, S296 – L2449 [again, see Fig. 4A]). Perhaps most interesting in this analysis is the finding that five of the nine S. buchneri SSU rDNA introns (S114, S287, S1052, S1210, S1506) are nested inside, or sister to introns at heterologous sites in the Lecanoromycetidae (see Figs. 3 and 4). This suggests that the five potentially ancestral SSU rDNA introns that were present in S. buchneri may have spread into 5 novel sites (i.e., S303, S1199, S1046, S497, S1516, respectively) during fungal evolution resulting in 10 different intron lineages. The timing of these intron movements suggests that most have occurred in derived members of the Lecanoromyectidae, in particular the lichenized Physicaceae [12]. The S1052 to S1046 movement, for example, must have been relatively recent because, despite extensive sampling [12,23], the novel S1046 intron is known only from the closely related taxa, Gymnoderma coccocarpum, Physcia stellaris, Lecanora dispersa, and Cladonia spp. (Lecanorales). The S114 to S303 movement is also likely to be recent because the derived S303 intron is limited to two members of the Physciaceae (Buellia georgei and Rinodina cacuminum). Other intron movements that are independent of S. buchneri include the S934 – L1025 and S296 – L2449 (see Figs. 3 and 4A,4B) sites. These groupings are only weakly supported in Fig. 3. All of the phylogenetic analyses with the 51-sequence data set suggest however that the Acarospora cf. dissipata S934 group I intron shares a specific evolutionary relationship with the intron at the L1025 site in a member of the Lecanoromycetidae (Porpidia albocaerulescens) and the Ostropomycetidae (Coccotrema pocillarium [Fig. 4A]). The majority-rule Bayesian consensus tree inferred from the intron core regions is consistent with these results showing that the intron transpositions are supported even when a restricted number of sites from the most highly conserved regions of the ribozymes are used in the analysis (Fig. 4B). This result argues against the spurious clustering of introns in Figures 3 and 4A resulting solely from long-branch attraction of divergent sequences.
Taken together, our phylogenetic analyses provide evidence not only for the ancient origin of many fungal introns that can be traced back to their presence in S. buchneri, but also provide indirect evidence for the spread of some of these introns into novel sites, thereby giving rise to additional vertically inherited sequences. Given this hypothesis, we predict that some taxa may maintain introns at both the ancestral and novel sites. This prediction is met for the S114 – S303 (Rinodina capensis), S287 – S1199 (Physconia perisidiosa, R. cacuminum), and S1506 – S1516 introns (e.g., P. perisidiosa). The low number of taxa containing both ancestral and derived introns suggests again that intron loss is widespread in fungal rDNA because in most cases one or both of the ribozymes have been lost from the genes. We cannot yet identify, however, the source of many introns that are putatively of a putative recent origin but are unrelated to the introns in Symbiotaphrina (e.g., S1389, L800, L1090). Clearly, external sources such as viruses or bacteria may also play an important role in the introduction of group I introns in fungal rDNA.
Reverse splicing group I intron spread
We suggest that reverse splicing may be an important mechanism of group I intron spread into distant rDNA sites in fungal nuclear rDNA for the following reasons: 1) Virtually none of the Pezizomycotina rDNA group I introns (including all of those implicated here in movement) contain endonucleases [10,26,27]. If endonucleases (encoded within the intron sequences) did mediate the lateral transfer of rDNA introns, then one would have to postulate complete loss of these coding regions after the introns had attained their present distribution. When endonucleases have been found in nuclear group I introns [see [10]], they appear to have been inserted into existing intron sequences and generally do not contain extensive deletions but rather frame-shift mutations or short truncations that result in their inactivation [e.g., [26,28], Reeb et al. unpublished data]. No large insertions (i.e., >200 nt) have been found in these introns that could be the remnant of an inactivated endonuclease open reading frame. 2) Generally, intron homing requires extensive exon flanking sequence identity (15 – 20 nt [2,7,8]) that is not found among the different rDNA group I intron movements (Fig. 4C). 3) In the cases of intron transposition shown here, all of the heterologous rDNA sites show high conservation of only 4 – 8 nt of the 5' exon flanking sequence (all exon sequences are from S. buchneri rDNA). This region is involved in the internal guide sequence interaction that is required for both forward- and reverse splicing (Fig. 1). In our analyses, the S287 – S1199 and S1046 – S1052 introns share the 5' flanking exon sequences 5'UAACRGGU3' and 5'UGKKGGU3', respectively (Fig. 4C). A close evolutionary relationship between the S497 – S1210 introns is also supported by this analysis (5'CCCU3'). This pattern of sequence conservation is predicted for reverse splicing-mediated intron spread [13,15]. A recent analysis of the S956 twin-ribozyme in the myxomycete Didymium iridis demonstrates that it can accurately reverse-splice into the homologous site in both E. coli and yeast rRNA [29]. Surprisingly, this intron does not integrate into related heterologous rRNA sites as has been reported for the T. thermophila ribozyme, which was shown to partially reverse-splice into 69 sites and completely integrate into one site in E. coli large subunit rRNA [16]. The S934 – L1025 intron group also displays a conserved motif at the site of insertion (5'CACCAC3') but one of these introns appears to have "slid" along the exon by one nucleotide either at the time of reverse splicing or sometime thereafter.
Conclusion
Given the results of our exhaustive study, we propose that reverse splicing is potentially an important mechanism of intron spread in Pezizomycotina nuclear rDNA. Our analyses provide strong support for the idea that the evolutionary history of nuclear group I introns may differentiate itself markedly from organellar group I introns which appear to rely primarily on homing for spread and may be characterized by massive and independent invasions into the homologous DNA sites of related taxa [e.g., [10,30]]. In contrast, the nuclear introns often have protracted histories in the host genomes [e.g., [12,23,25,31]] and their spread into novel sites may be a direct consequence of this long-term stability; i.e., existing introns reverse-splice at different times into heterologous sites. The availability of reverse transcriptase in the cell and the fixation rate of intron-containing alleles in the multi-copy rDNA gene family are likely the primary determinants of the rate of reverse splicing-mediated movement. Limited analyses of intron+ and intron- strains fail to show any clear advantage or disadvantage to the host cell when introns are present [32,33]. This suggests that mobile group I introns may be "silent" parasites that have no measurable phenotype [34,35]. The neutral nature of group I introns suggests that their spread and loss may be stochastic events with rare movement occurring through reverse splicing and the more frequent intron loss occurring most likely through chance or when intron mobility, splicing, and/or processing (e.g., degradation) poses a cost to the host cell [7]. Successful group I intron excision is of critical importance because these sequences are normally found in functionally important regions of rDNA [36,37].
Our study suggests that the beetle gut symbiont S. buchneri [see [38]] contains a set of introns that has likely been vertically inherited in later-diverging fungi with reverse-splicing spread of some of these ribozymes into ectopic rDNA sites, particularly in the intron-rich Lecanoromycetes. Our data do not support the idea that S. buchneri is a vector for facilitating intron spread in the fungi. We did not find an unexpectedly close phylogenetic relationship between any of the introns in this taxon with others in our data set, which would have indicated a recent later transfer event. The phylogenetic positions of the S. buchneri introns relative to other ribozymes (when bootstrap and/or Bayesian support is found) are generally consistent with the expectations of host relationships. However, these group I intron data do not have enough resolving power to allow us to address the possibility of more ancient transfer events involving S. buchneri. In conclusion, it is apparent from our work that nuclear group I introns in the fungi have remarkably complex evolutionary histories, therefore our analyses are likely only scratching the surface of intron movement in these taxa. More detailed studies of specific intron lineages [e.g., [11,23,25]] will more accurately reveal the dynamics of fungal rDNA group I intron evolution.
Methods
Fungal cultures, DNA extraction, and PCR amplification
Five cultures of Symbiotaphrina spp. were acquired for this study: S. buchneri (JCM9740) and S. kochii (JCM9739, CBS250.77, CBS588.63, CBS589.63). DNA was isolated from these cultures using the Puregene Kit (Gentra Systems) following the manufacturer's protocol for filamentous fungi. The nuclear small (SSU) and large subunit (LSU) rDNA were amplified for each strain by PCR. In addition we amplified the RPB2 (second-largest subunit of RNA polymerase II) nuclear gene from S. buchneri JCM9740 and S. kochii CBS589.63. PCR products were purified from agarose gels using GELase (Epicenter) and directly sequenced on both strands using an ABI PRISM 3700 DNA Analyzer (Applied Biosystems), and Big Dye (Perkin-Elmer, Applied Biosystems). The PCR primers used in this study came from various sources [39-47] or were designed specifically for Symbiotaphrina spp. The rDNA and RPB2 sequences for Symbiotaphrina spp. are available from GenBank under the following accession numbers respectively: S. buchneri JCM9740 (rDNA = DQ248313; RPB2 = DQ248315), S. kochii CBS589.63 (rDNA = DQ248314; RPB2 = DQ248316).
Sequence alignments and phylogenetic analyses
Fungal host phylogeny
To provide a framework for understanding group I intron evolution in the fungi, we reconstructed a phylogeny of the Pezizomycotina that included Symbiotaphrina spp. This tree was inferred from a combined DNA data set of nuclear SSU rDNA, LSU rDNA, and RPB2 DNA (2100 nt) from 84 ascomycetes with one basidiomycete as the outgroup. These data are available from GenBank. The combined data set was analyzed using a GTR + Γ + I model of evolution for each of the five data partitions (SSU, LSU, RPB2-1st, -2nd, -3rd codon positions). Bayesian analysis (MrBayes V3.0b4 [48]) was initiated using a random tree from the combined dataset with four chains running simultaneously for 5,000,000 generations, and trees sampled every 100 generations. The first 10,000 trees were discarded (burnin) and a majority rule consensus tree was generated from the remaining 40,000 (post burnin) trees. A neighbor-joining analysis was also used to calculate bootstrap support values for nodes in the Bayesian consensus tree. The supra-generic taxon names used in this tree follow [49].
rDNA group I intron phylogeny
The ten group I introns found in Symbiotaphrina spp. were aligned with 179 fungal group I introns at 28 different rDNA sites. The non-Symbiotaphrina introns are published [e.g., [12]] and available either from GenBank or from the Comparative RNA Web Site [50]. The introns from the 28 rDNA sites represent well the diversity of fungal rDNA group I introns although we excluded a small number of introns from other sites that were either difficult to align or to unambiguously identify their rDNA genic position (e.g., S940, S1049, S1201). The 189 group I introns were aligned through juxtaposition of the secondary structural elements P1–P9 found in nuclear group I introns [31,51,52]. For this procedure we used, wherever possible, existing secondary structures from representatives of different intron insertion sites [e.g., [10,11,53-55]] to guide the alignment. We did not attempt to include all available fungal group I introns (there are nearly 1200 group I introns in this group [see [50]]) but sampled (given the taxonomic distribution) evenly the different lineages. Our approach was designed to provide an overall view of fungal group I intron phylogeny and is not expected to detect lateral transfers within intron lineages that would be apparent in detailed analyses of introns at particular rDNA sites and the host phylogeny [e.g., [10,11,20,23,25,55]]. Given the large number of introns and rDNA genic sites to consider, we divided the phylogenetic analyses into increasingly more focused data sets. The initial data set of 189 introns was used to gain broad insights into group I intron phylogeny and in particular, the distribution of the S. buchneri introns within the tree. This tree provided evidence for the vertical evolution and movement of introns. Thereafter, we reduced the data set to a representative group of 116 introns to increase the phylogenetic resolution. Finally, we included the putative reverse splicing candidates in data sets of 51 and 34 sequences. The sequences were pruned approximately uniformly from the trees to retain the diversity of introns at the different rDNA sites. This approach was necessary to gain meaningful insights into group I intron evolution because phylogenetic methods often perform poorly under the situations used here; i.e., the interrelationships of many divergent lineages need to be resolved with a relatively small data set.
A total of 136 aligned positions were selected for the initial phylogenetic analyses (alignment available from DB upon request). For these data, we used two different approaches to infer the phylogeny. First, we used the single parameter Jukes-Cantor (JC) evolutionary model [56] with neighbor-joining (NJ) tree reconstruction to estimate a tree. This "simple" model is potentially useful for large data sets with short (in this case, highly divergent) sequences when multiple parameter estimates are expected to have high associated variances [e.g., [57,58]]. Under such conditions, the maximum likelihood method may give an incorrect topology [59]. Branch lengths will, however, be underestimated under the JC model. The JC-NJ tree was inferred using PAUP* (V4.0b10 [60]) and bootstrap analyses (2000 replications) were done to assess the support for monophyletic groups in the JC-NJ tree. In the second approach, we used the parameter-rich GTR + Γ model ([61] i.e., estimated proportion of invariant sites = 0.0147) in a Bayesian inference as described above to calculate posterior probabilities for nodes in the intron tree. In this analysis, a random starting tree was initiated and run for 3,000,000 generations with trees sampled every 1000th generation. To increase the probability of chain convergence, the first 2,000 trees were discarded as burnin and the remaining 1,000 were used to calculate the posterior probabilities.
Based on the analysis of the 189-sequence data set, we generated a second reduced intron data set of 116 sequences that maintained the diversity of intron sites in the large data set. A JC-NJ tree was inferred from these data (with bootstrap support values) and Bayesian posterior probabilities were calculated for the tree as described above. In addition, we did an unweighted maximum parsimony (MP) bootstrap analysis of the data. For this method, a heuristic search was used with each of the 2000 bootstrap pseudosamples and starting trees were obtained using random additions (10 rounds) with tree bisection-reconnection branch swapping. The 51-sequence data set (138 nt in this alignment) was analyzed with the JC-NJ, MP, and Bayesian methods as described above. In addition, we did a maximum likelihood bootstrap analysis of these data. In this approach, the gamma value (with 4 rate categories) and the transition/transversion ratio were estimated using PAUP*. Bootstrap analyses (100 replicates) were then done using DNAML (PHYLIP V3.6b [62]) with 1 random taxon addition and global rearrangements. We also generated a second reduced alignment of 34 introns that included only the catalytic core (66 nt) of the fungal group I introns [11]. Analysis of the core region alignment allowed us to assess whether the most highly conserved region of these ribozymes resulted in essentially the same tree as when the more variable regions were included. For the core alignment, we used Bayesian inference (as described above) to infer a 50% majority-rule consensus phylogeny from the final 1000 trees in the posterior distribution. In all of these intron phylogenies, the evolutionarily distantly related group IE introns [5,37] were used to root the subtree of IC introns [10]. Finally, we used the maximum likelihood-based Shimodaira-Hasegawa statistical test [63] to assess likelihood support alternative intron topologies.
Authors' contributions
VR and FL did the sequencing of group I introns and did the phylogenetic analysis of the fungal host tree and contributed to the manuscript. DMS sequenced fungal introns and contributed to the manuscript. DB conceived of and supervised this study and wrote the manuscript. All authors read and approved the final manuscript.
Acknowledgements
This work was supported by grants awarded to D. B. from the National Science Foundation (MCB 0110252, DEB 0107754) and Avis E. Cone and Stanley fellowships from the University of Iowa to D. S. We also thank Peik Haugen (Iowa, now in Tromso, Norway) for helpful discussions and two anonymous reviewers for their constructive criticisms.
==== Refs
Cech TR Self-splicing RNA: implications for evolution Int Rev Cytol 1985 93 3 22 3891660
Lambowitz AM Belfort M Introns as mobile genetic elements Annu Rev Biochem 1993 62 587 622 8352597 10.1146/annurev.bi.62.070193.003103
Bhattacharya D The origin and evolution of protist group I introns Protist 1998 149 113 122
Haugen P Simon D Bhattacharya D The natural history of group I introns Trends Genet 2005 21 111 119 15661357 10.1016/j.tig.2004.12.007
Bhattacharya D Cannone JJ Gutell RR Group I intron lateral transfer between red and brown algal ribosomal RNA Curr Genet 2001 40 82 90 11570520 10.1007/s002940100227
Chevalier BS Stoddard BL Homing endonucleases: structural and functional insight into the catalysts of intron/intein mobility Nucleic Acids Res 2001 29 3757 3774 11557808 10.1093/nar/29.18.3757
Dujon B Group I introns as mobile genetic elements: facts and mechanistic speculations-a review Gene 1989 82 91 114 2555264 10.1016/0378-1119(89)90034-6
Belfort M Perlman PS Mechanisms of intron mobility J Biol Chem 1995 270 30237 30240 8530436 10.1074/jbc.270.51.30237
Goddard MR Burt A Recurrent invasion and extinction of a selfish gene Proc Natl Acad Sci USA 1999 96 13880 13885 10570167 10.1073/pnas.96.24.13880
Haugen P Reeb V Lutzoni F Bhattacharya D The evolution of homing endonuclease genes and group I introns in nuclear rDNA Mol Biol Evol 2004 21 129 140 14595099 10.1093/molbev/msh005
Haugen P Runge HJ Bhattacharya D Long-term evolution of the S788 fungal nuclear small subunit rRNA group I introns RNA 2004 10 1084 1096 15208444 10.1261/rna.5202704
Bhattacharya D Friedl T Helms G Vertical evolution and intragenic spread of lichen-fungal group I introns J Mol Evol 2002 55 74 84 12165844 10.1007/s00239-001-2305-x
Woodson SA Cech TR Reverse self-splicing of the Tetrahymena group I intron: implication for the directionality of splicing and for intron transposition Cell 1989 57 335 345 2702692 10.1016/0092-8674(89)90971-9
Roman J Woodson SA Reverse splicing of the Tetrahymena IVS: evidence for multiple reaction sites in the 23S rRNA RNA 1995 1 478 490 7489509
Roman J Woodson SA Integration of the Tetrahymena group I intron into bacterial rRNA by reverse splicing in vivo Proc Natl Acad Sci USA 1998 95 2134 2139 9482851 10.1073/pnas.95.5.2134
Roman J Rubin MN Woodson SA Sequence specificity of in vivo reverse splicing of the Tetrahymena group I intron RNA 1999 5 1 13 9917062 10.1017/S1355838299981244
Lynch M Conery JS The origins of genome complexity Science 2001 302 1401 1404 14631042 10.1126/science.1089370
Noda H Kodama K Phylogenetic position of the yeastlike endosymbionts of anobiid beetles Appl Environ Microbiol 1996 62 162 167 8572692
Lonergan KM Gray MW The ribosomal RNA gene region in Acanthamoeba castellanii mitochondrial DNA A case of evolutionary transfer of introns between mitochondria and plastids? J Mol Biol 1994 239 476 499 8006963 10.1006/jmbi.1994.1390
Hibbett DS Phylogenetic evidence for horizontal transmission of group I introns in the nuclear ribosomal DNA of mushroom-forming fungi Mol Biol Evol 1996 13 903 917 8751999
Cho Y Qiu YL Kuhlman P Palmer JD Explosive invasion of plant mitochondria by a group I intron Proc Natl Acad Sci USA 1998 95 14244 14249 9826685 10.1073/pnas.95.24.14244
Friedl T Besendahl A Pfeiffer P Bhattacharya D The distribution of group I introns in lichen algae suggests that lichenization facilitates intron lateral transfer Mol Phylogenet Evol 2000 14 342 352 10712840 10.1006/mpev.1999.0711
Simon DM Moline J Helms G Friedl T Bhattacharya D Divergent histories of rDNA group I introns in the lichen family Physciaceae J Mol Evol 2005 60 434 446 15883879 10.1007/s00239-004-0152-2
Jones KG Blackwell M Ribosomal DNA sequence analysis places the yeast-like genus Symbiotaphrina within filamentous ascomycetes Mycologia 1996 88 212 218
Nikoh N Fukatsu T Evolutionary dynamics of multiple group I introns in nuclear ribosomal RNA genes of endoparasitic fungi of the genus Cordyceps Mol Biol Evol 2001 18 1631 1642 11504844
Johansen S Haugen P A complex group I intron in Nectria galligena rDNA Microbiology 1999 145 516 517 10217484
Perotto S Nepote-Fus P Saletta L Band C Young JP A diverse population of introns in the nuclear ribosomal genes of ericoid mycorrhizal fungi includes elements with sequence similarity to endonuclease-coding genes Mol Biol Evol 2000 17 44 59 10666705
Haugen P Huss VA Nielsen H Johansen S Complex group-I introns in nuclear SSU rDNA of red and green algae: evidence of homing-endonuclease pseudogenes in the Bangiophyceae Curr Genet 36 345 353 10654088 10.1007/s002940050509
Birgisdottir AB Johansen S Site-specific reverse splicing of a HEG-containing group I intron in ribosomal DNA Nucleic Acids Res 2005 33 2042 2051 15817568 10.1093/nar/gki341
Palmer JD Adams KL Cho Y Parkinson CL Qiu YL Song K Dynamic evolution of plant mitochondrial genomes: mobile genes and introns and highly variable mutation rates Proc Natl Acad Sci U S A 2000 97 6960 6966 10860957 10.1073/pnas.97.13.6960
Bhattacharya D Surek B Ruesing M Melkonian M Group I introns are inherited through common ancestry in the nuclear-encoded rRNA of Zygnematales (Charophyceae) Proc Natl Acad Sci USA 1994 91 9916 9920 7937917
Nielsen H Engberg J Sequence comparison of the rDNA introns from six different species of Tetrahymena Nucleic Acids Res 1985 13 7445 7455 4059059
Johanningmeier U Heiss S Construction of a Chlamydomonas reinhardtii mutant with an intronless psbA gene Plant Mol Biol 1993 22 91 99 8499620 10.1007/BF00038998
Belfort M Phage T4 introns: self-splicing and mobility Annu Rev Genet 1990 24 363 385 2088173 10.1146/annurev.ge.24.120190.002051
Cech TR Self-splicing of group I introns Annu Rev Biochem 1990 59 543 568 2197983 10.1146/annurev.bi.59.070190.002551
Turmel M Cote V Otis C Mercier JP Gray MW Lonergan KM Lemieux C Evolutionary transfer of ORF-containing group I introns between different subcellular compartments (chloroplast and mitochondrion) Mol Biol Evol 1995 12 533 545 7659010
Bhattacharya D Simon D Huang J Cannone JJ Gutell RR The exon context and distribution of rRNA introns BMC Evol Biol 2003 3 7 12716459 10.1186/1471-2148-3-7
Suh S-O McHugh JV Pollock DD Blackwell M The beetle gut: a hyperdiverse source of novel yeasts Mycol Res 2005 109 261 265 15912941 10.1017/S0953756205002388
Gardes M Bruns TD TS primers with enhanced specificity for basidiomycetes-application to the identification of mycorrhizae and rusts Mol Ecol 1993 2 113 118 8180733
Gargas A Taylor JW Polymerase chain reaction (PCR) primers for amplifying and sequencing 18S rDNA from lichenized fungi Mycologia 1992 84 589 592
Guadet J Julien J Lafay JF Brygoo Y Phylogeny of some Fusarium species, as determined by large-subunit rRNA sequence comparison Mol Biol Evol 1989 6 227 242 2622333
Kauff F Lutzoni F Phylogeny of the Gyalectales and Ostropales (Ascomycota, Fungi): among and within order relationships based on nuclear ribosomal RNA small and large subunits Mol Phylogenet Evo 2002 25 138 156 10.1016/S1055-7903(02)00214-2
Miadlikowska J McCune B Lutzoni F Pseudocyphellaria perpetua, a new lichen from western North America Bryologist 2002 105 1 10
Reeb V Lutzoni F Roux C Contribution of RPB2 to multilocus phylogenetic studies of the euascomycetes (Pezizomycotina, Fungi) with special emphasis on the lichen-forming Acarosporaceae and evolution of polyspory Mol Phylogenet Evol 2004 32 1036 1060 15288074 10.1016/j.ympev.2004.04.012
Spatafora JW Mitchell TG Vilgalys R Analysis of genes coding for small-subunit rRNA sequences in studying phylogenetics of dematiaceous fungal pathogens J Clin Microbiol 1995 33 1322 1326 7615749
Vilgalys R Hester M Rapid genetic identification and mapping of enzymatically amplified ribosomal DNA from several Cryptococcus species J Bacteriol 1990 172 4238 4246 2376561
White TJ Bruns T Lee S Taylor J Innes MA, Gelfand DH, Sninsky JJ, White TJ Amplification and direct sequencing of fungal ribosomal RNA gene for phylogenetics PCR Protocols: a Guide to Methods and Applications 1990 San Diego: Academic Press 315 322
Ronquist F Huelsenbeck JP MrBayes 3: Bayesian phylogenetic inference under mixed models Bioinformatics 2003 19 1572 1574 12912839 10.1093/bioinformatics/btg180
Lutzoni F Kauff F Cox CJ McLaughlin D Celio G Dentinger B Padamsee M Hibbett D James TY Baloch E Grube M Reeb V Hofstetter V Schoch C Arnold AE Miadlikowska J Spatafora J Johnson D Hambleton S Crockett M Shoemaker R Sung G-H Lücking R Lumbsch T O'Donnell K Binder M Diederich P Ertz D Gueidan C Hall B Hansen K Harris RC Hosaka K Lim Y-W Liu Y Matheny B Nishida H Pfister D Rogers J Rossman A Schmitt I Sipman H Stone J Sugiyama J Yahr R Vilgalys R Assembling the fungal tree of life: progress, classification, and evolution of subcellular traits Amer J Bot 2004 91 1446 1480
Cannone JJ Subramanian S Schnare MN Collett JR D'Souza LM Du Y Feng B Lin N Madabusi LV Müller KM Pande N Shang Z Yu N Gutell RR The Comparative RNA Web (CRW) Site: an online database of comparative sequence and structure information for ribosomal, intron, and other RNAs BMC Bioinformatics 2002 3 1 31 11818024 10.1186/1471-2105-3-1
Michel F Westhof E Modelling of the three-dimensional architecture of group I catalytic introns based on comparative sequence analysis J Mol Biol 1990 216 585 610 2258934 10.1016/0022-2836(90)90386-Z
Golden BL Gooding AR Podell ER Cech TR A preorganized active site in the crystal structure of the Tetrahymena ribozyme Science 1998 282 259 264 9841391 10.1126/science.282.5387.259
DePriest PT Been MD Numerous group I introns with variable distributions in the ribosomal DNA of a lichen fungus J Mol Biol 1992 228 315 321 1453441 10.1016/0022-2836(92)90819-6
Suh SO Jones KG Blackwell M A group I intron in the nuclear small subunit rRNA gene of Cryptendoxyla hypophloia, an ascomycetous fungus: evidence for a new major class of group I introns J Mol Evol 1999 48 493 500 10198116
Coates BS Hellmich RL Lewis LC Nuclear small subunit rRNA group I intron variation among Beauveria spp provide tools for strain identification and evidence of horizontal transfer Curr Genet 2002 41 414 424 12228811 10.1007/s00294-002-0317-8
Jukes TH Cantor CR Munro HN Evolution of protein molecules Mammalian Protein Metabolism 1969 New York: Academic Press 21 132
Bruno WJ Halpern AL Topological bias and inconsistency of maximum likelihood using wrong models Mol Biol Evol 1999 16 564 566 10331281
Piontkivska H Efficiencies of maximum likelihood methods of phylogenetic inferences when different substitution models are used Mol Phylogenet Evol 2004 31 865 873 15120384 10.1016/j.ympev.2003.10.011
Nei M Kumar S Takahashi K The optimization principle in phylogenetic analysis tends to give incorrect topologies when the number of nucleotides or amino acids used is small Proc Natl Acad Sci USA 1998 95 12390 12397 9770497 10.1073/pnas.95.21.12390
Swofford DL PAUP*: Phylogenetic Analysis Using Parsimony (* and Other Methods) 40b10 2003 Sunderland: Sinauer
Rodriguez F Oliver JF Marin A Medina JR The general stochastic model of nucleotide substitutions J Theor Biol 1990 142 485 501 2338834
Felsenstein J PHYLIP, V363 2004 Seattle: Department of Genome Sciences, University of Washington
Shimodaira H Hasegawa M Multiple comparisons of log-likelihoods with applications to phylogenetic inference Mol Biol Evol 1999 16 1114 1116
|
16300679
|
PMC1299323
|
CC BY
|
2021-01-04 16:29:17
|
no
|
BMC Evol Biol. 2005 Nov 21; 5:68
|
utf-8
|
BMC Evol Biol
| 2,005 |
10.1186/1471-2148-5-68
|
oa_comm
|
==== Front
BMC GastroenterolBMC Gastroenterology1471-230XBioMed Central London 1471-230X-5-361628750510.1186/1471-230X-5-36Case ReportFibrinogen storage disease without hypofibrinogenemia associated with estrogen therapy Simsek Z [email protected] O [email protected] M [email protected] T [email protected] B [email protected] G [email protected] S [email protected] Gazi University Faculty of Medicine, Gastroenterology Department, Ankara, Turkey2 Gazi University Faculty of Medicine, Pathology Department, Ankara, Turkey2005 15 11 2005 5 36 36 21 2 2005 15 11 2005 Copyright © 2005 Simsek et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Cytoplasmic inclusion bodies within hepatocytes may have different etiologies, including the Endoplasmic Reticulum Storage Diseases (ERSDs). ERSD is a pathological condition characterized by abnormal accumulation of proteins destined for secretion in the endoplasmic reticulum of hepatocytes; it may be congenital (primary) or acquired (secondary). Fibrinogen storage disease is a form of ERSD.
Case Presentation
We present a case of fibrinogen storage disease secondary to estrogen replacement therapy. Its causal relationship to the drug is shown by histological, immunohistochemical and ultrastructural studies of paired liver biopsies obtained during and after the drug therapy.
Conclusion
The liver biopsies of patients with idiopathic liver enzyme abnormalities should be carefully evaluated for cytoplasmic inclusion bodies and, although rare, fibrinogen deposits.
==== Body
Background
Cytoplasmic inclusions in hepatocytes are encountered in a wide range of clinical conditions including chronic hepatitis B, α1-antitrypsin deficiency, type IV glycogenosis, Lafora's disease, fibrinogen storage disease, and in drug reactions [1,2]. Endoplasmic Reticulum Storage Disease (ERSD) of the liver is a pathological condition characterized by abnormal accumulation of proteins destined for secretion in the endoplasmic reticulum. The term ERSD encompasses α1-antitrypsin deficiency, the prototype disorder, plus fibrinogen storage disease and α1-antichymotrypsin deficiency [3]. ERSD can be congenital or acquired. The congenital form is associated with the accumulation of only one protein, due to its altered structure. The accumulation is permanent and affects mainly the rough endoplasmic reticulum (RER). The most common cause of this condition is α1-antitrypsin deficiency. Acquired [secondary] form occurs due to exogenous agents or in the presence of concomitant diseases such as infections, and therefore structurally normal proteins accumulate in several cellular organelles.
In this article, we report a case of fibrinogen storage disease of the liver secondary to the use of oral contraceptives. This is the first reported case of fibrinogen storage disease secondary to estrogen ingestion. Informed consent was obtained from the patient regarding her recruitment in this case study.
Case presentation
A forty-year-old female was referred to our department for evaluation of elevated liver enzymes in May, 2004. She had no systemic symptoms. A year before, she had had a total abdominal hysterectomy and bilateral salpingo-oophorectomy. Since then, she had been on hormone replacement therapy with estrogen. Six months ago, during her routine follow-up, elevation of transaminase levels was noted. This elevation persisted for the next 6 months. She was then referred to our department for further investigation.
Laboratory features of the patient regarding this period are shown in Table 1. Detailed questioning of the patient revealed no history of past alcohol abuse. Screening for viral antigens and for antibodies against them, as well as screening for viral nucleotide-acid sequences (including HCV-RNA/-cDNA PCR), found no evidence of infection. Levels of auto-antibodies, serum transferrin saturation, serum ferritin, serum ceruloplasmin, serum copper, serum α1-antitrypsin, and 24-hour urinary copper excretion were either negative or within normal limits. Abdominal ultrasonography was unremarkable. Doppler ultrasound imaging of portal and hepatic veins and the hepatic artery revealed no thrombosis.
A liver biopsy was performed. Histopathological examination revealed focal necrosis and mild fibrotic activity. Macrovesicular steatosis affecting 5% of hepatocytes, in addition to several foci of prominent nuclear pleomorphism and hyperchromasia were also noted. However, the most remarkable finding was the presence of abundant intracytoplasmic, sharply bordered, pale eosinophilic inclusions that exhibited a ground-glass appearance (Figure 1). Periodic acid-Schiff (PAS), diastase-treated PAS and colloidal iron stains were applied. Immunohistochemistry for hepatitis B surface (mouse monoclonal antibody clone: Tg, Neomarkers) and core (mouse monoclonal antibody clone: Tordji-22, Signet) antigens, α1-antitrypsin (rabbit antibody, Zymed), fibrinogen (rabbit anti-human antibody, Dako), α fetoprotein (rabbit polyclonal antibody, Signet), complement 3 (mouse monoclonal antibody clone: HAV 3–4, Dako) and complement 4 (rabbit anti-human antibody, Dako) were performed. A three-step streptavidin biotin method using AEC as chromogen was applied in the immunohistochemical procedure. The inclusions were negative to PAS and colloidal iron. No reactivity with antibodies against hepatitis antigens, α1-antitrypsin and complement 3 was observed. Diffuse and strong positivity was observed for fibrinogen in cytoplasmic inclusions (Figure 2). We also detected less positivity for complement 4 and α-fetoprotein.
Paraffin-embedded tissue was processed for electron microscopic examination. Ultrastructurally, the inclusions were found to be moderately electron-dense, finely granular, homogenous bodies that are encircled by membranes, corresponding to a dilated endoplasmic reticulum (Figure 3). There were no parallel arrays of tubular structures in the inclusions.
Serum fibrinogen level was within normal limits. We performed tests for qualitative fibrinogen function. Prothrombine time, partial thromboplastin time and thrombin time were also within normal limits. The patient had no history of prolonged bleeding episodes, menorrhagia, or epistaxis. Since infectious, auto-immune or metabolic diseases were excluded, a drug effect was suspected. Therefore, estrogen therapy was ceased and the patient was followed monthly in order to check transaminase levels. Her serum transaminase levels decreased gradually at each monthly visit and normalized at the third month. Then we decided to perform a control liver biopsy. Parenchymal and portal inflammation, steatosis and necroinflammatory activity were remarkably reduced in the second liver biopsy (Figure 4). Again, the same staining procedures were applied. Compared to the first biopsy, fibrinogen positive inclusions decreased dramatically to the point where only one hepatocyte contained an inclusion. Fibrotic activity and nuclear dysplastic findings remained almost the same.
Conclusion
The patient had elevated transaminase levels without any accompanying clinical symptoms. Her first liver biopsy revealed cytoplasmic inclusion bodies within hepatocytes associated with chronic hepatitis and mild fibrotic activity. The inclusions were sharply bordered and faintly eosinophilic together with a ground-glass appearance. They were negative to PAS and colloidal iron; therefore type IV glycogenoses and Lafora's disease were ruled out on morphological grounds, respectively. No staining with hepatitis antigens and α1-antitrypsin was observed immunohistochemically. Diffuse and strong positivity for fibrinogen was observed. Her serum fibrinogen levels, in addition to prothrombin, partial thromboplastin and thrombin times – excluding dysfibrinogenemia – were within normal limits. The patient was not the offspring of a consanguineous marriage.
Fibrinogen storage in hepatocytes has been previously reported in patients with or without hypofibrinogenemia. The disease was first shown in German families as a familial hypofibrinogenemia [4,5]. Later, specific mutations in hereditary cases were reported [6,7]. Callea et al., placed fibrinogen storage disorders in the ERSD group along with α1-antitrypsin and α1-antichymotrypsin deficiencies, and proposed a classification of fibrinogen storage diseases based on morphologic and clinical evidence, and thus divided the entity into three types [3]. In type I, the inclusions are defined as round or polygonal, weakly eosinophilic cytoplasmic deposits with irregular borders. This type is a hereditary hypofibrinogenemia, genetically characterized by the presence of mutant variants of the fibrinogen molecule, namely fibrinogen Brescia and fibrinogen Aguadilla [3,6,7]. The other two types are rarer. Type II inclusions are large, hyaline bodies that occupy the entire cytoplasm and result in a ground-glass appearance, while round, eosinophilic globules surrounded by a clear halo constitute the inclusions of type III. At the ultrastructural level, type I inclusions are characterized by the presence of densely packed tubular structures, arranged in curved bundles resulting in a finger-print-like appearance. Type II inclusions appear as granular or fragmented fibrillar material and type III inclusions contain a central core resembling the tubular structures of Type I and a surrounding similar to that of Type II. In all types, the inclusions correspond to dilated cisternae of RER.
Cases of fibrinogen storage disease without hypofibrinogenemia have been recently reported. One such patient had atypical type of cytoplasmic inclusions and was considered to demonstrate a congenital case [8]. Two other patients revealed fibrinogen storage in the liver during an acute infectious episode together with Type II inclusions. The latter were thought to have transient, acquired forms of the disease [9].
The inclusions observed in our patient were large, hyaline bodies occupying the entire cytoplasm, resulting in a ground-glass appearance. This morphology is consistent with Type II inclusions. Neither the characteristics of Type I inclusions, such as irregular borders, a clear halo surrounding the inclusion or an acicular shape, nor the findings consistent with Type III inclusions [round, eosinophilic globules with a clear halo around] were present in our patient. Electron microscopy revealed homogenous, granular material within dilated ER, which are also the characteristics of Type II inclusions. Tubular structures found in Type I and partially in Type III inclusions were not present either. These findings allowed us to consider the patient's inclusions to be Type II.
In the first biopsy, fibrinogen storage was accompanied with accumulation of complement 4 and α fetoprotein. This simultaneous deposition of multiple proteins destined for secretion indicates a dysfunction in the secretory apparatus, rather than the retention of a single mutant protein [10]. Callea et al., claimed that this finding may help distinguish between the acquired and congenital forms of the disease [3]. A single, genetically abnormal protein cannot be translocated from RER to smooth endoplasmic reticulum (SER), therefore it accumulates in RER [11]. This type of selective and exclusive retention of fibrinogen defines the congenital and the more frequent form of the disease. On the other hand, structurally normal proteins can get trapped in RER, SER and/or secretory vesicles due to a secretory dysfunction, which might be caused by exogenous agents like alcohol, and by drugs such as colchicine [3]. Thus in our case, a drug-induced, acquired form of fibrinogen storage is further supported by the above findings as well as the dramatic resolution of histopathologic and biochemical parameters after cessation of estrogen administration.
We ascribed the findings of chronic hepatitis and mild fibrotic activity in our patient to fibrinogen storage. Indeed, cases without hypofibrinogenemia, as well as those with hereditary hypofibrinogenemia and those that have a mutational basis, were shown to develop chronic hepatitis, fibrosis or cirrhosis, in the literature [3,4,8,9]. Importantly, low levels of fibrinogen does not necessarily accompany the hepatic storage of fibrinogen, which was the case in our patient. This may also support the secondary, acquired nature of the disease.
Relying on the results explained above, we conclude that, in our patient an oral regimen of estrogen is associated with fibrinogen storage disorder. To our knowledge, this is the first case reported to date to show hepatic fibrinogen storage secondary to estrogen ingestion. The liver biopsies of patients with idiopathic liver enzyme abnormalities should be carefully evaluated for cytoplasmic inclusion bodies and, although rare, fibrinogen deposits.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Figures and Tables
Figure 1 Intracytoplasmic inclusions with ground-glass appearance affecting most of the hepatocytes in this area. Mononuclear inflammatory infiltrate in a portal area is also seen. Hematoxylin-eosin, ×40. Inset: Higher magnification of this area, (H&E, ×200).
Figure 2 Strong and diffuse immunohistochemical positivity of the inclusions for fibrinogen, ×40. İnset: Same area with higher magnification. (×200).
Figure 3 On ultrastructure, inclusions appear as finely granular, homogeneous material bounded by membranes (arrows). ×4400
Figure 4 In the second biopsy, inclusions and chronic hepatitic changes are dramatically reduced, (H&E, ×200).
Table 1 Laboratory features of the patient with exogenous estrogen
Laboratory parameter Result Normal range
Aspartate aminotransferase (IU/L) 565 0–40
Alanine aminotransferase (IU/L) 545 0–40
Lactate dehydrogenase (IU/L) 424 254–474
Alkaline phosphatase (IU/L) 354 112–330
gamma-glutamyltranspeptidase (IU/L) 98 0–50
Total bilirubin (mg/dl) 0.5 0.2–1.2
Blood urea nitrogen (mg/dl) 18 2–20
Creatinine (mg/dl) 0.4 0.4–1.0
Triglycerides (mg/dl) 154 29–180
Total cholesterol (mg/dl) 180 128–200
Blood glucose (mg/dl) 85 60–110
Blood ammonium-N (mg/dl) 11 40–85
==== Refs
Hashimoto K Hoshii Y Takahashi M Mitsuno S Hanai N Watanabe Y Ishihara T Use of monoclonal antibody against Lafora bodies for the immunocytochemical study of ground-glass inclusions in hepatocytes due to cyanamide Histopathology 2001 39 60 65 11454045 10.1046/j.1365-2559.2001.01127.x
Vazquez JJ Ground-glass hepatocytes: light and electron microscopy. Characterization of the different types Histol Histopathol 1990 5 379 386 1966881
Callea F Brisigotti M Fabbretti G Bonino F Desmet VJ Hepatic endoplasmic reticulum storage diseases Liver 1992 12 357 362 1470006
Wehinger H Klineg O Alexandrakis E Schumann J Witt J Seydewitz HH Hereditary hypofibrinogenaemia with fibrinogen storage in the liver Eur J Pediatr 1983 141 109 112 6662138 10.1007/BF00496800
Pfeifer U Ormanns W Klinge O Hepatocellular fibrinogen storage in familial hypofibrinogenemia Virchows Arch B Cell Pathol Incl Mol Pathol 1981 36 247 55 6116338
Brennan SO Wyatt J Medicina D Callea F George PM Fibrinogen Brescia: hepatic endoplasmic reticulum storage and hypofibrogenaemia because of a gamma 284 Gly Arg mutation Am J Pathol 2000 157 189 196 10880389
Brennan SO Maghzal G Shneider BL Gordon R Magid MS George PM Novel fibrinogen gamma375 Arg-->Trp mutation (fibrinogen aguadilla) causes hepatic endoplasmic reticulum storage and hypofibrinogenemia Hepatology 2002 36 652 8 12198657 10.1053/jhep.2002.35063
Abukawa D Tazawa Y Noro T Nakagawa M Iinuma K Sugiyama K Knisely AS Cytoplasmic inclusion bodies and minimal hepatitis: fibrinogen storage without hypo-fibrogenaemia Pediatr Dev Pathol 2001 4 304 309 11370269 10.1007/s100240010174
Marucci G Morandi L Macchia S Betts CM Tardio ML Monte Dal PR Pession A Foschini MP Fibrinogen storage disease without hypofibrinogenemia associated with acute infection Histopathology 2003 42 22 5 12493021 10.1046/j.1365-2559.2003.01551.x
Ng IOL Ng M Lai ECS Wu PC Endoplasmic storage disease of liver: characterization of intracytoplasmic hyaline inclusions Histopathology 1989 15 473 81 2480934
Medicina D Fabbretti G Brennan SO George PM Kudryk B Callea F Genetic and immunological characterization of fibrinogen inclusion bodies in patients with hepatic fibrinogen storage and liver disease Ann N Y Acad Sci 2001 936 522 5 11460509
|
16287505
|
PMC1299324
|
CC BY
|
2021-01-04 16:03:27
|
no
|
BMC Gastroenterol. 2005 Nov 15; 5:36
|
utf-8
|
BMC Gastroenterol
| 2,005 |
10.1186/1471-230X-5-36
|
oa_comm
|
==== Front
BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-1611628865810.1186/1471-2164-6-161Research ArticleConstruction and characterization of a genomic BAC library for the Mus m. musculus mouse subspecies (PWD/Ph inbred strain) Jansa Petr [email protected] Petr [email protected] Jiří [email protected] Institute of Molecular Genetics, Academy of Sciences of the Czech Republic and Center for Applied Genomics, Vídeňská 1083, CZ-142 20, Prague 4, Czech Republic2005 16 11 2005 6 161 161 16 8 2005 16 11 2005 Copyright © 2005 Jansa et al; licensee BioMed Central Ltd.2005Jansa et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The genome of classical laboratory strains of mice is an artificial mosaic of genomes originated from several mouse subspecies with predominant representation (>90%) of the Mus m. domesticus component. Mice of another subspecies, East European/Asian Mus m. musculus, can interbreed with the classical laboratory strains to generate hybrids with unprecedented phenotypic and genotypic variations. To study these variations in depth we prepared the first genomic large insert BAC library from an inbred strain derived purely from the Mus m. musculus-subspecies. The library will be used to seek and characterize genomic sequences controlling specific monogenic and polygenic complex traits, including modifiers of dominant and recessive mutations.
Results
A representative mouse genomic BAC library was derived from a female mouse of the PWD/Ph inbred strain of Mus m. musculus subspecies. The library consists of 144 768 primary clones from which 97% contain an insert of 120 kb average size. The library represents an equivalent of 6.7 × mouse haploid genome, as estimated from the total number of clones carrying genomic DNA inserts and from the average insert size. The clones were arrayed in duplicates onto eight high-density membranes that were screened with seven single-copy gene probes. The individual probes identified four to eleven positive clones, corresponding to 6.9-fold coverage of the mouse genome. Eighty-seven BAC-ends of PWD/Ph clones were sequenced, edited, and aligned with mouse C57BL/6J (B6) genome. Seventy-three BAC-ends displayed unique hits on B6 genome and their alignment revealed 0.92 single nucleotide polymorphisms (SNPs) per 100 bp. Insertions and deletions represented 0.3% of the BAC end sequences.
Conclusion
Analysis of the novel genomic library for the PWD/Ph inbred strain demonstrated coverage of almost seven mouse genome equivalents and a capability to recover clones for specific regions of PWD/Ph genome. The single nucleotide polymorphism between the strains PWD/Ph and C57BL/6J was 0.92/100 bp, a value significantly higher than between classical laboratory strains. The library will serve as a resource for dissecting the phenotypic and genotypic variations between mice of the Mus m. musculus subspecies and classical laboratory mouse strains.
==== Body
Background
PWD/Ph is a highly inbred strain currently at 81 generations of brother × sister matings. It originated from the Mus m. musculus mouse subspecies by systematic inbreeding of a pair of wild mice trapped in 1972 [1,2]. The mouse subspecies M. m. musculus and M. m. domesticus diverged from their common ancestor about 300 thousand years [3] to 1 million years ago [4] and at present they display signs of incomplete reproductive isolation [5-7]. As a consequence of the interrupted gene flow between both subspecies, the mice of the PWD/Ph strain exhibit a high degree of DNA polymorphisms and a broad range of phenotypic differences when compared to classical laboratory strains [2,8]. Because of this unique feature, the PWD/Ph inbred strain has been nominated among 15 mouse strains, the genomes of which are being resequenced using high-density oligonucleotide array technology by Perlegen Sciences, Inc. [9]. Moreover, PWD/Ph serves as the chromosome donor strain in construction of a set of C57BL/6-ChrPWD chromosome substitution strains (Gregorova, Forejt and coworkers, in preparation).
Bacterial Artificial Chromosome (BAC) genomic libraries are source of large genomic DNA insert clones for sequencing projects, physical mapping and isolation of intact genes [10,11]. Although BAC clones may carry large inserts of genomic DNA (up to 200 kb) they display low rate of de novo rearrangements and are easy to handle. These features are in strong favor of the BAC libraries over the Yeast Artificial Chromosome (YAC) libraries, which can contain up to 60% of chimeric clones [12]. Transgenic mice can be generated using BAC clones to examine candidate genes in context of all regulatory DNA elements required for their function and the phenotype of a mutant mouse can be rescued by BAC transgenesis [13,14]. Moreover a targeted modification at exact positions within a genomic BAC clone can be introduced by recombineering [15,16].
Here we report construction and characterization of the PWD/Ph BAC library, the first genomic library of the Mus m. musculus mouse subspecies. This library together with the upcoming panel of chromosome substitution strains will serve as a tool for analysis of complex traits by taking advantage of the evolutionary divergence between the two closely related mouse subspecies.
Results and discussion
Construction of the PWD/Ph-BAC library
The BAC library was prepared by cloning the EcoRI-partially digested genomic DNA from the spleen of a PWD/Ph female mouse in the vector pBACe3.6. Female DNA was chosen to gain an unbiased representation of the X chromosome in the library. The primary clones were picked and arrayed in 377 individual 384-well plates. The library consists of two segments containing 54 144 and 90 624 clones, respectively. Together 144 768 primary clones were arrayed on eight high-density nylon membranes (18 342 clones in doublets per membrane). The high-density membranes were utilized in subsequent hybridization experiments.
Average insert size of the library
The average insert size of the library was determined on a set of 400 randomly selected BAC clones. DNA samples were prepared from 164 and 236 BAC clones from the library segments 1 and 2, respectively, and subjected to NotI restriction analysis. The products of the digestion reactions were resolved by pulsed-field gel electrophoresis (PFGE) along with the high molecular weight markers. The average insert size for the first and the second library segment was 101.1 kb (SD ± 21.4) and 129.5 kb (SD ± 14.7), respectively (Figure 1). In the first and second library segments 6% and 1.2% clones were observed without insert, corresponding to 97% of insert-containing clones for the entire library. Estimation of 6.7-fold redundancy of the library was based on the average insert size (120 kb) and 2.6 × 109 bp size of the mouse genome.
Figure 1 Insert size distribution in two segments of the PWD/Ph BAC library. The segment 1 (□) represents 37.4% of the clones and its average insert size was 101.1 kb (SD ± 21.4). The segment 2 (■) represents 62.6% of the clones and its average insert size was 129.5 kb (SD ± 14.7). The average insert size of the entire library was 120 kb.
Library screening and BAC end sequencing
A probability to find any given unique sequence in the library is 99.85%, according to the published formula (P = 1 - eN.ln(1-I/GS), where P is probability, N is number of clones, I is insert size, and GS is size of genome) [17]. To further assess the genome redundancy and possible cloning bias of the library experimentally we performed a screening of the library with 7 single-copy gene probes. The probes were designed to amplify PCR products on the PWD/Ph genomic DNA template (Table 1). Seven probes detected in total 48 positive clones by hybridization on 8 high-density library membranes, 4 to 11 clones for each individual probe. The average number of clones recognized by a single probe was 6.9, in good accordance with the assessment of the library redundancy based on the average insert size.
Table 1 Hybridization of single-copy gene probes on high-density membrane
Gene/Primer GenBank accession/Primer sequence Genomic position1 Length Positive clones
Mash NM_008554
mMashCgi-F ACCCGGTTCCTCGCGAGCACTTTTC chr7:130,673,330 358 bp 5/48
mMashCgi-B AGCGCAGCGTCTCCACCTTACTCAG chr7:130,672,998
Adseverin NM_009132
CpG-Ads-1F TCTTGGAGGGTCATACTCATT chr12:35,153,275 516 bp 7/48
CpG-Ads-1R GCAGCTCAAAATAATTACGAC chr12:35,152,780
Igf2r NM_010515
Igf2r-H4F TCAGAACACTGGTGAGCAGTGGG chr17:12,150,732 244 bp 11/48
Igf2-H4R GAGGGTAGGATTCCGTTGCAAGG chr17:12,150,509
Tbp NM_013684
* probe derived from tbp-1942 clone chr17:14,324,440 1085 bp 4/48
chr17:14,334,524
Usp26 NM_031388
Usp26-A AATGTAACGAAGGGAGAAGTG chrX:44,101,298 206 bp 3/48
Usp26-B AGGCTTTGCCTTCTTATCGAG chrX:44,101,113
Xist AY618354
mXistF AGTGGGTGTTCAGGGCGTGG chrX:94,885,925 293 bp 11/48
mXistR CTATCCCCTAGTCCTCTGCGG chrX:94,885,652
Tex13 NM_031381
Tex13 pub-1F ACCAGAGTTGGGAACAACTAA chrX:130,816,501 220 bp 7/48
Tex13 pds-1R CTGTTGTAGAGGGTAGAGGTT chrX:130,816,302
1 mm5 assembly, UCSC
To characterize the inserts of the PWD/Ph BAC library at the DNA sequence level we sequenced and manually edited 87 BAC ends from 47 BAC clones (total 38,339 nucleotides). The BAC end sequences (BESs) were masked for repeats and aligned on the C57BL/6J mouse genome. BES pairs of 29 BAC clones mapped to unique positions in the B6 genome on the opposite DNA strands within the distance up to 200 kb (Additional file 1). The mapping allowed us to estimate the average insert size of the BAC clones based on their locations on the B6 genome as 127 kb, which was slightly higher estimate than the average insert size acquired by restriction analysis (120 kb). These values corresponded well with the average insert size calculated for another set of clones recovered by the library screening described above (Table 2). A BES pair belonging to the clone 307-9O mapped to two distinct chromosomes. Whether it represents a chimeric insert or a chromosomal rearrangement in the PWD/Ph genome remains to be determined by fluorescence in situ hybridization (FISH) analysis. For each of the additional 13 BAC clones we found unambiguous positions for only one BES of a pair. Mapping of remaining 14 BESs was prevented by a high content of repetitive elements.
Table 2 BAC end sequences of the positive clones mapped on the C57BL/6J genome
Gene Probe position BAC clone/primer BES position Strand Insert size (bp)
mapped PFGE
Mash chr7:130,672,998 327-5I/T7 chr7:130,587,753 + 132,317 145,000
327-5I/SP6 chr7:130,720,069 -
Ads chr12:35,152,780 262-11G/SP6 multiple hits n.d. 105,000
262-11G/T7 chr12:35,192,443 -
266-2E/SP6 chr12:35,090,782 + 109,399 115,000
266-2E/T7 chr12:35,200,180 -
Igf2r chr17:12,150,509 279-5I/SP6 chr17:12,057,702 + 125,000 120,000
279-5I/T7 chr17:12,182,701 -
282-5D/T7 chr17:12,028,446 + 144,525 145,000
282-5D/SP6 chr17:12,172,970 -
245-11P/SP6 chr17:12,000,500 + 186,822 195,000
245-11P/T7 chr17:12,187,321 -
Tbp chr17:14,324,440 297-2I/T7 chr17:14,291,556 + 147,925 145,000
297-2I/SP6 chr17:14,439,480 -
Usp26 chrX:44,101,113 293-21P/T7 chrX:44,086,068 + 184,994 190,000
293-21P/SP6 chrX:44,271,061 -
Xist chrX:94,885,652 269-9H/SP6 chrX:94,817,775 + 150,314 160,000
269-9H/T7 chrX:94,968,088 -
255-3L/T7 chrX:94,756,622 + 134,724 145,000
255-3L/SP6 chrX:94,891,345 -
271-12L/SP6 chrX:94,854,496 + 131,966 140,000
271-12L/T7 chrX:94,986,461 -
Tex13 chrX:130,816,302 257-22B/SP6 chrX:130,761,904 + 124,211 120,000
257-22B/T7 chrX:130,886,114 -
322-4J/SP6 chrX:130,761,839 + 117,691 110,000
322-4J/T7 chrX:130,879,529 -
BAC ends (BESs) of the positive clones identified on the membranes no. 6 and no. 7 were sequenced, mapped and aligned to the mouse genome (mm5 assembly, UCSC). The coordinates on the mouse genome are shown for the probes as well as for the ends of genomic inserts. An average insert size 140.8 kb was calculated from mapped positions and 141.2 kb was determined by pulsed-field gel electrophoresis (PFGE).
Analysis of SNPs and DNA polymorphism
To find out the degree of nucleotide polymorphism between the PWD/Ph and C57BL/6J mouse strains, we aligned 73 uniquely mapped BESs (32,182 nucleotides) with their C57BL/6J genomic counterparts and found 297 single nucleotide substitutions. The calculated SNP rate 0.92 per 100 bp is significantly higher than SNP frequency between laboratory strains [18-20] and corresponds well to the rate between the closely related subspecies Mus m. molossinus and the C57BL/6J strain (0.96%) [21]. The insertions and deletions (indels) were found with lower frequency than SNPs: single nucleotide indels occurred with frequency 0.19% while multinucleotide indels with only 0.08% frequency. All nucleotide changes observed in the alignments of 87 PWD/Ph BESs and their B6 counterparts are summarized in Additional files 2 and 3. The high number of SNPs of the PWD/Ph strain is reflected by a high frequency of genetic and phenotypic variations between PWD/Ph and B6 inbred mice. An initial study performed to compare behavior of the PWD/Ph inbred strain with the B6 revealed substantial behavioral differences between these two strains [8]. Using dense SNP maps of various laboratory and wild-derived inbred strains [20,22] it will be possible to map genes responsible for particular complex traits more efficiently. For ultimate validation of candidate genes genomic BAC libraries will be highly desirable.
Conclusion
The first genomic BAC library was constructed for the Mus m. musculus subspecies of the house mouse, represented by the PWD/Ph inbred strain. The quality of the PWD BAC library was verified by hybridization with seven unique probes that identified multiple positive clones. BAC end sequencing provided a new piece of evidence on the high incidence of SNPs (0.92/100 bp) between C57BL/6J and PWD/Ph inbred strains. The mouse PWD/Ph BAC library will serve as a tool for functional genomics of complex genetic traits with the ultimate goal to identify and clone responsible genes. The PWD BAC library will become accessible to the scientific community via RZPD, Berlin, Germany [23].
Methods
Mouse strain and DNA isolation
Mouse manipulation was in accordance with the Czech Animal Protection Act No. 246/92, 162/93, and decrees No. 311/97, fully compatible with the NIH Publication No. 85-23, revised 1985. A female mouse of the PWD/Ph inbred strain, derived from the Mus mus musculus subspecies [2] was used for high molecular weight DNA (HMW-DNA) preparation. The mouse was killed by cervical dislocation, spleen dissected and single cell suspension prepared in PBS using a glass homogenizer. The agarose-embeded HMW-DNA was prepared as described in detail elsewhere [24].
Library construction
The agarose HMW-DNA plugs were subjected to pre-electrophoresis in a CHEF-DR-III apparatus (BioRad) in 1% agarose and 0.5 × TBE buffer for 12 hrs (4 V/cm, 5 s pulse, 14°C). Genomic DNA was partially digested with the mixture of EcoRI endonuclease and EcoRI methylase. The optimal ratio of the enzymes was determined by titration: usually 5–25 units of methylase per 1 unit of endonuclease were employed. DNA fragments were prepared by slight modification of an approach described before [24]. Briefly, DNA fragments were separated from the digested agarose plugs in the CHEF-DR-III in 1% agarose and 0.5 × TBE buffer for 16 hrs (5 V/cm, 0.1 to 40 s pulse, 14°C). Subsequently, three stripes corresponding to fragment size between 150 kb and 200 kb were excised and subjected to another size selection by additional electrophoresis in 0.5 × TBE buffer for 12 hr (5 V/cm, 2.5 to 4.5 s pulse, 14°C). The second size selection effectively removed short fragments while keeping long fragments in the agarose strips. The appropriate fragments were isolated by electroelution and ligated to the EcoRI site of the pBACe3.6 vector [25]. The ligation mixtures were dialyzed on ice in a well created by 0.5% agarose with 1 M glucose for 1 hr. The desalted ligation mixtures were electroporated into E. coli electrocompetent DH10B ElectroMax cells (Invitrogen) by a Gene Pulser apparatus (BioRad) in 0.1 cm cuvette with the following parameters: voltage 1.8 V, impedance 200 Ω, capacitance 25 μF, time constant between 3.5 to 4.5 s. The electroporated cells were diluted in 1 ml SOC medium and incubated in an orbital shaker at 37°C and 200 rpm for 1 hr. The titer of each electroporation reaction was determined by spreading an aliquot on selection agar plates (LB, 20 μg/ml chloramphenicol, 5% sucrose) as described [24]. The remainder containing the primary clones was supplemented with glycerol to the final concentration of 15%, then quickly frozen in liquid nitrogen and stored at -80°C. The frozen stocks of the primary clones were recovered and spread on large selection plates. The colonies were picked with multi-functional robotical system Gene TacTM-G3 (Genomic Solutions) and arrayed in 377 individual 384-well dishes containing LB medium supplied by 7.5% glycerol and 20 μg/ml chloramphenicol. The clones were gridded using the robot on 8 nylon membranes (18 342 unique clones in duplicates per membrane). Afterwards, the bacterial colonies were lysed, their DNA denatured, and crosslinked to the membranes by standard methods [26].
Estimation of average insert size
One hundred sixty-two clones from the library segment 1 and 236 clones from the segment 2 were randomly picked and grown in 15 ml 2xYT medium (16 g/L tryptone, 10 g/L yeast extract, 5 g/L NaCl, 20 μg/ml chloramphenicol) at 37°C and 300 rpm for 18–20 h. BAC DNA was prepared using a modification of the standard protocol based on alkalic lysis. Briefly, 15 ml of the overnight culture was spun down, the bacterial pellet was resuspended in 300 μl of lysis buffer I (50 mM glucose, 25 mM Tris-HCl with pH = 8.0, 10 mM EDTA), then lysed with 600 μl of freshly made lysis buffer II (0.2 M NaOH, 1% SDS) and precipitated with 450 μl of lysis buffer III (3 M KOAc, pH = 4.8) followed by incubation on ice for 1 hour. The resulted precipitate was spun down in a microfuge for 15 min at maximum speed. The BAC DNA was further precipitated at room temperature with 0.6 volumes of isopropanol for 30 min and centrifuged at maximum speed for 10 min. The pellet was washed with 70% ethanol, air dried shortly and dissolved in 25 μl of TE. The BAC DNA was subjected to NotI (Fermentas) restriction overnight to achieve complete digestion. The reactions were resolved along with the mid range PFG marker I (New England Biolabs, cat # N3551S) in the CHEF-DR-III in 1% agarose and 0.5 × TBE buffer for 16 hrs (5 V/cm, 0.1 to 40 s pulse, 14°C). The insert size was estimated after ethidium bromide staining and the average insert size for both segments of the library was calculated.
Hybridization of high-density colony filters
Seven probes for single-copy mouse genes were used to screen the library on high-density membranes. Six of them were produced by PCR on 50 ng of HMW-DNA isolated from the brain of a female PWD/Ph mouse. The primers were designed using Oligo 6 (MBI) software and the GenBank mRNA sequences (Table 1). PCR was performed at standard conditions: 96°C for 1 min, 30 cycles at 96°C for 10 s, 58°C for 20 s, 72°C for 1 min, and final extension at 72°C for 3 min. The 1085-bp DNA fragment of Tbp (U63933) was derived from the tbp-1942 cDNA clone (kind gift from Dr. Trachtulec) [27]. The resulted amplicons and Tbp fragment were purified and labeled by random priming with [α-32P]dCTP. Hybridization was done in Church buffer [28] with 15% formamide at 60°C over night. The membranes were washed in 0.2 × SSC, 0.1% SDS buffer at 42°C for 20 min and then in 0.2 × SSC, 0.1% SDS buffer at 60°C for 10 min. The membranes were then autoradiographed for 2 to 10 days, the positive clones identified and recovered from the frozen 384-well plates.
BAC end sequencing and analysis
BAC DNA was prepared from 60 clones as described above and purified using a QIAGEN kit following the manufacturer's instructions. The sequencing was performed using a Big Dye Terminator v3.1 cycle sequencing kit in an ABI 310 instrument (Applied Biosystems) with primers T7 (GGTCGAGCTTGACATTGTAG) and SP6 (GATCCTCCCGAATTGACTAGTG). Each DNA sample was sequenced twice. The sequences from the same BAC end were aligned and manually edited in order to obtain a consensus sequence. BESs were masked for mouse repeats using RepeatMasker [29] (sensitive settings) and aligned to the mouse genome sequence (mm5 assembly, May 2004, UCSC) [30,31] using BLAT [32]. The mouse genome sequence had already been soft-masked for repeats by UCSC and BLAT was set to produce all possible alignments (tile size = 10, minimum score = 0, minimum sequence identity = 0). The hits were filtered to keep only those with minimum alignment ratio = 0.8. After manual inspection, a list of BESs mapped to unambiguous positions in the genome was compiled. The corresponding genomic sequences were excised and aligned with the appropriate BESs (unmasked sequence) using SSEARCH [33] (standard settings). A Perl script was written to process the pair-wise alignments and enumerate the sequence polymorphisms (SNPs, insertions, deletions, etc). The visualization of DNA polymorphisms was made by TeXshade LaTeX package [34]. All intermediate steps were performed using customized Perl scripts and utilities available from UCSC website [35].
Authors' contributions
PJ constructed the library, participated in the sequence alignment and drafted the manuscript with JF and PD. PD carried out the sequence alignments and the analysis of DNA polymorphism. JF conceived the study, participated in its design and coordination. PJ and JF wrote the manuscript and all authors read and approved its final version.
Supplementary Material
Additional file 1
BAC end sequences mapped on C57BL/6 mouse genome.
Click here for file
Additional file 2
Summary of polymorphisms detected in 73 BAC end sequences.
Click here for file
Additional file 3
Polymorphisms detected in 73 BAC end sequences.
Click here for file
Acknowledgements
We thank Heinz Himmelbauer for invaluable help at the start of the project, Jaroslav Doležel for the access to the core facility of his laboratory, Jaroslav Janda, Michaela Libošvárová and Radka Tusková for help with picking and gridding the library, Vladana Fotopulosová for help with DNA sequencing, and Zdeněk Trachtulec and Šárka Takáčová for reading the manuscript. This work was supported by the Ministry of Education, Youth and Sports of the Czech Republic Grant No. 1M6837805002 – Center for Applied Genomics, by the Grant Agency of the Academy of Sciences Grant AVOZ50520514 and by NIH grant 1R01 HG00318. Jiří Forejt was supported as an International Scholar of the Howard Hughes Medical Institute by Grant 55000306.
==== Refs
Pavljukova H Forejt J Private communication Mouse News Lett 1981 65 44
Gregorova S Forejt J PWD/Ph and PWK/Ph inbred mouse strains of Mus m. musculus subspecies--a valuable resource of phenotypic variations and genomic polymorphisms Folia Biol (Praha) 2000 46 31 41 10730880
Boursot P Din W Anand R Darviche D Dod B VonDeimling F Talwar GP Bonhomme F Origin and radiation of the house mouse: Mitochondrial DNA phylogeny. Journal of Evolutionary Biology 1996 9, 391-415. 391 415 10.1046/j.1420-9101.1996.9040391.x
Silver LM Mouse Genetics Concepts and Applications Concepts and Applications 1995 , Oxford University Press 376
Forejt J Ivanyi P Genetic studies on male sterility of hybrids between laboratory and wild mice (Mus musculus L.) Genet Res 1974 24 189 206 4452481
Forejt J Hybrid sterility in the mouse Trends Genet 1996 12 412 417 8909138 10.1016/0168-9525(96)10040-8
Storchova R Gregorova S Buckiova D Kyselova V Divina P Forejt J Genetic analysis of X-linked hybrid sterility in the house mouse Mamm Genome 2004 15 515 524 15366371
Fernandes C Liu L Paya-Cano JL Gregorova S Forejt J Schalkwyk LC Behavioral Characterization of Wild Derived Male Mice ( Mus musculus musculus ) of the PWD/Ph Inbred Strain: High Exploration Compared to C57BL/6J Behav Genet 2004 34 621 630 15520518 10.1007/s10519-004-5589-y
Perlegen Sciences, Inc.
Osoegawa K Zhu B Shu CL Ren T Cao Q Vessere GM Lutz MM Jensen-Seaman MI Zhao S de Jong PJ BAC resources for the rat genome project Genome Res 2004 14 780 785 15060022 10.1101/gr.2033904
McPherson JD Marra M Hillier L Waterston RH Chinwalla A Wallis J Sekhon M Wylie K Mardis ER Wilson RK Fulton R Kucaba TA Wagner-McPherson C Barbazuk WB Gregory SG Humphray SJ French L Evans RS Bethel G Whittaker A Holden JL McCann OT Dunham A Soderlund C Scott CE Bentley DR Schuler G Chen HC Jang W Green ED Idol JR Maduro VV Montgomery KT Lee E Miller A Emerling S Kucherlapati Gibbs R Scherer S Gorrell JH Sodergren E Clerc-Blankenburg K Tabor P Naylor S Garcia D de Jong PJ Catanese JJ Nowak N Osoegawa K Qin S Rowen L Madan A Dors M Hood L Trask B Friedman C Massa H Cheung VG Kirsch IR Reid T Yonescu R Weissenbach J Bruls T Heilig R Branscomb E Olsen A Doggett N Cheng JF Hawkins T Myers RM Shang J Ramirez L Schmutz J Velasquez O Dixon K Stone NE Cox DR Haussler D Kent WJ Furey T Rogic S Kennedy S Jones S Rosenthal A Wen G Schilhabel M Gloeckner G Nyakatura G Siebert R Schlegelberger B Korenberg J Chen XN Fujiyama A Hattori M Toyoda A Yada T Park HS Sakaki Y Shimizu N Asakawa S Kawasaki K Sasaki T Shintani A Shimizu A Shibuya K Kudoh J Minoshima S Ramser J Seranski P Hoff C Poustka A Reinhardt R Lehrach H A physical map of the human genome Nature 2001 409 934 941 11237014 10.1038/35057157
Banfi S Ledbetter SA Chinault AC Zoghbi HY An easy and rapid method for the detection of chimeric yeast artificial chromosome clones Nucleic Acids Res 1992 20 1814 1579482
Antoch MP Song EJ Chang AM Vitaterna MH Zhao Y Wilsbacher LD Sangoram AM King DP Pinto LH Takahashi JS Functional identification of the mouse circadian Clock gene by transgenic BAC rescue Cell 1997 89 655 667 9160756 10.1016/S0092-8674(00)80246-9
Abe K Hazama M Katoh H Yamamura K Suzuki M Establishment of an efficient BAC transgenesis protocol and its application to functional characterization of the mouse Brachyury locus Exp Anim 2004 53 311 320 15297704 10.1538/expanim.53.311
Lee EC Yu D Martinez de Velasco J Tessarollo L Swing DA Court DL Jenkins NA Copeland NG A highly efficient Escherichia coli-based chromosome engineering system adapted for recombinogenic targeting and subcloning of BAC DNA Genomics 2001 73 56 65 11352566 10.1006/geno.2000.6451
Gong S Yang XW Li C Heintz N Highly efficient modification of bacterial artificial chromosomes (BACs) using novel shuttle vectors containing the R6Kgamma origin of replication Genome Res 2002 12 1992 1998 12466304 10.1101/gr.476202
Clarke L Carbon J A colony bank containing synthetic Col El hybrid plasmids representative of the entire E. coli genome Cell 1976 9 91 99 788919 10.1016/0092-8674(76)90055-6
Lindblad-Toh K Winchester E Daly MJ Wang DG Hirschhorn JN Laviolette JP Ardlie K Reich DE Robinson E Sklar P Shah N Thomas D Fan JB Gingeras T Warrington J Patil N Hudson TJ Lander ES Large-scale discovery and genotyping of single-nucleotide polymorphisms in the mouse Nat Genet 2000 24 381 386 10742102 10.1038/74215
Wiltshire T Pletcher MT Batalov S Barnes SW Tarantino LM Cooke MP Wu H Smylie K Santrosyan A Copeland NG Jenkins NA Kalush F Mural RJ Glynne RJ Kay SA Adams MD Fletcher CF Genome-wide single-nucleotide polymorphism analysis defines haplotype patterns in mouse Proc Natl Acad Sci U S A 2003 100 3380 3385 12612341 10.1073/pnas.0130101100
Frazer KA Wade CM Hinds DA Patil N Cox DR Daly MJ Segmental phylogenetic relationships of inbred mouse strains revealed by fine-scale analysis of sequence variation across 4.6 mb of mouse genome Genome Res 2004 14 1493 1500 15289472 10.1101/gr.2627804
Abe K Noguchi H Tagawa K Yuzuriha M Toyoda A Kojima T Ezawa K Saitou N Hattori M Sakaki Y Moriwaki K Shiroishi T Contribution of Asian mouse subspecies Mus musculus molossinus to genomic constitution of strain C57BL/6J, as defined by BAC-end sequence-SNP analysis Genome Res 2004 14 2439 2447 15574823 10.1101/gr.2899304
Pletcher MT McClurg P Batalov S Su AI Barnes SW Lagler E Korstanje R Wang X Nusskern D Bogue MA Mural RJ Paigen B Wiltshire T Use of a dense single nucleotide polymorphism map for in silico mapping in the mouse PLoS Biol 2004 2 e393 15534693 10.1371/journal.pbio.0020393
RZPD, Berlin, Germany
Osoegawa K de Jong PJ Frengen E Ioannou PA Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, Smith JA and Struhl K Construction of Bacterial Arteficial Chromosome (BAC/PAC) Librarries Current Protocols in Molecular Biology 2001 1 , John Wiley & Sons, Inc 5.9.1 5.9.33
Frengen E Weichenhan D Zhao B Osoegawa K van Geel M de Jong PJ A modular, positive selection bacterial artificial chromosome vector with multiple cloning sites Genomics 1999 58 250 253 10373322 10.1006/geno.1998.5693
Ausubel FM Brent R Kingston RE Moore DD Seidman JG Smith JA Struhl K Chanda VB Current Protocols in Molecular Biology 2005 , John Wiley & Sons, Inc.
Trachtulec Z Hamvas RM Forejt J Lehrach HR Vincek V Klein J Linkage of TATA-binding protein and proteasome subunit C5 genes in mice and humans reveals synteny conserved between mammals and invertebrates Genomics 1997 44 1 7 9286694 10.1006/geno.1997.4839
Church GM Gilbert W Genomic sequencing Proc Natl Acad Sci U S A 1984 81 1991 1995 6326095
Smit AF Hubley R Green P RepeatMasker Open-3.0
Karolchik D Baertsch R Diekhans M Furey TS Hinrichs A Lu YT Roskin KM Schwartz M Sugnet CW Thomas DJ Weber RJ Haussler D Kent WJ The UCSC Genome Browser Database Nucleic Acids Res 2003 31 51 54 12519945 10.1093/nar/gkg129
Waterston RH Lindblad-Toh K Birney E Rogers J Abril JF Agarwal P Agarwala R Ainscough R Alexandersson M An P Antonarakis SE Attwood J Baertsch R Bailey J Barlow K Beck S Berry E Birren B Bloom T Bork P Botcherby M Bray N Brent MR Brown DG Brown SD Bult C Burton J Butler J Campbell RD Carninci P Cawley S Chiaromonte F Chinwalla AT Church DM Clamp M Clee C Collins FS Cook LL Copley RR Coulson A Couronne O Cuff J Curwen V Cutts T Daly M David R Davies J Delehaunty KD Deri J Dermitzakis ET Dewey C Dickens NJ Diekhans M Dodge S Dubchak I Dunn DM Eddy SR Elnitski L Emes RD Eswara P Eyras E Felsenfeld A Fewell GA Flicek P Foley K Frankel WN Fulton LA Fulton RS Furey TS Gage D Gibbs RA Glusman G Gnerre S Goldman N Goodstadt L Grafham D Graves TA Green ED Gregory S Guigo R Guyer M Hardison RC Haussler D Hayashizaki Y Hillier LW Hinrichs A Hlavina W Holzer T Hsu F Hua A Hubbard T Hunt A Jackson I Jaffe DB Johnson LS Jones M Jones TA Joy A Kamal M Karlsson EK Karolchik D Kasprzyk A Kawai J Keibler E Kells C Kent WJ Kirby A Kolbe DL Korf I Kucherlapati RS Kulbokas EJ Kulp D Landers T Leger JP Leonard S Letunic I Levine R Li J Li M Lloyd C Lucas S Ma B Maglott DR Mardis ER Matthews L Mauceli E Mayer JH McCarthy M McCombie WR McLaren S McLay K McPherson JD Meldrim J Meredith B Mesirov JP Miller W Miner TL Mongin E Montgomery KT Morgan M Mott R Mullikin JC Muzny DM Nash WE Nelson JO Nhan MN Nicol R Ning Z Nusbaum C O'Connor MJ Okazaki Y Oliver K Overton-Larty E Pachter L Parra G Pepin KH Peterson J Pevzner P Plumb R Pohl CS Poliakov A Ponce TC Ponting CP Potter S Quail M Reymond A Roe BA Roskin KM Rubin EM Rust AG Santos R Sapojnikov V Schultz B Schultz J Schwartz MS Schwartz S Scott C Seaman S Searle S Sharpe T Sheridan A Shownkeen R Sims S Singer JB Slater G Smit A Smith DR Spencer B Stabenau A Stange-Thomann N Sugnet C Suyama M Tesler G Thompson J Torrents D Trevaskis E Tromp J Ucla C Ureta-Vidal A Vinson JP Von Niederhausern AC Wade CM Wall M Weber RJ Weiss RB Wendl MC West AP Wetterstrand K Wheeler R Whelan S Wierzbowski J Willey D Williams S Wilson RK Winter E Worley KC Wyman D Yang S Yang SP Zdobnov EM Zody MC Lander ES Initial sequencing and comparative analysis of the mouse genome Nature 2002 420 520 562 12466850 10.1038/nature01262
Kent WJ BLAT--the BLAST-like alignment tool Genome Res 2002 12 656 664 11932250 10.1101/gr.229202. Article published online before March 2002
Pearson WR Flexible sequence similarity searching with the FASTA3 program package Methods Mol Biol 2000 132 185 219 10547837
Beitz E TEXshade: shading and labeling of multiple sequence alignments using LATEX2 epsilon Bioinformatics 2000 16 135 139 10842735 10.1093/bioinformatics/16.2.135
UCSC Genome Browser, Jim Kent's Web Page
|
16288658
|
PMC1299325
|
CC BY
|
2021-01-04 16:32:47
|
no
|
BMC Genomics. 2005 Nov 16; 6:161
|
utf-8
|
BMC Genomics
| 2,005 |
10.1186/1471-2164-6-161
|
oa_comm
|
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-1031628864910.1186/1471-2334-5-103Research Article"Bacterial Meningitis in children and adolescents: an observational study based on the national surveillance system" Dickinson Félix O [email protected]érez Antonio E [email protected] Department of epidemiology, Institute of Tropical Medicine "Pedro Kourí", Havana, Cuba2 Autopista Novia del Mediodía Km. 6 Municipio Lisa, Ciudad de La Habana, Cuba. P.O. Box 601 Marianao 13, Cuba2005 15 11 2005 5 103 103 24 12 2004 15 11 2005 Copyright © 2005 Dickinson and Pérez; licensee BioMed Central Ltd.2005Dickinson and Pérez; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Bacterial meningitis is a group of life threatening infections that mostly affect children and adolescents, and may be the cause of severe neurological sequelae. Cuba has implemented massive vaccination programmes against both Neisseria meningitidis (serogroup C in 1979 and B in 1987), and Haemophilus influenzae type b (1999), two of the main causal pathogens. We described and discussed some epidemiological aspects of the current status of bacterial meningitis to learn from the Cuban experience.
Methods
A nationwide observational study on children and adolescents from 1 to 18 years old was carried out from 1998 to 2003, estimating the incidence and case-fatality rate by age group and causal pathogens, as well as the seasonality and frequency of overcrowded dormitories. The association between disease and attendance to day care centres or boarding schools was estimated by using relative risk (Chi-squared test and Fisher Exact Test).
Results
The overall number of cases was 1023; the incidence ranged from 3.4 to 8.5 per 100 000 population, with the higher figures in children 1–5 years old (16.8 per 100 000 population). Streptococcus pneumoniae, Haemophilus influenzae type b and Neisseria meningitidis serogroup B were the main identified agents. The average case-fatality rate was 10.5% and the most lethal agents were Streptococcus pneumoniae (27%) and Haemophilus influenzae type b (10.7%). Overall percentage of cases who slept in overcrowded dormitories was 15%, reaching 30.6% in adolescents. Seasonality was only evident among meningococcal meningitis cases between September–October. The attendance to boarding high school showed an association with disease only in 1998 and 1999 (RR = 2.1; p > 0.05).
Conclusion
The highest incidence of bacterial meningitis was observed among children from 1–5 years old. Pneumococcus was both the leading causal and the most lethal agent. Sleeping in overcrowded dormitories was more frequent among adolescents. No strong association was observed between the bacterial meningitis and attendance to day care centres or boarding schools. The incidence of bacterial meningitis in Cuba is declining after massive vaccination programmes against Neisseria meningitidis serogroup B and C and Haemophilus influenzae type b through a national immunisation program.
==== Body
Background
Bacterial meningitis (BM) is a group of severe infections that frequently affect children and adolescents, carrying a high case-fatality rate (CFR). It may cause transient or permanent deafness as well as other severe neurological sequelae in survivors [1-3].
The World Health Organisation estimated that annually BM causes at least 1.2 million cases worldwide and of those 135 000 deaths. Though other bacteria may cause BM, Neisseria meningitidis, Haemophilus influenzae type b (Hib) and Streptococcus pneumoniae (Spn) are the main triad responsible for more than 80% of all cases. Gram negative bacteria, particularly E. coli, Streptococci (different from S. pneumoniae), Listeria monocitogenes and Staphylococci may also cause BM [3].
BM has been reported in Cuba since the beginning of the 20th century [4]. The mandatory report of this group of infections has been implemented since 1961 as part of national communicable diseases surveillance.
Following a meningococcal disease epidemic in 1979, an independent surveillance system was put into place [4]. After controlling the epidemic, other germs like Hib and Pneumococcus increased their aetiological importance. In 1997 public health authorities decided to extend the surveillance to all bacteria, as well as to improve epidemiological and microbiological data through the Bacterial Meningitis National Surveillance System (BMSS) [5]. This allowed a high level of accuracy and quality information to be collected which could be compared with high surveillance systems international standards [5-8].
On the other hand, Cuba is the only country in the world which has carried out massive vaccination initiatives against Neisseria meningitidis serogroup C (1979) and B (1987) [9], and later in 1999 against Hib [10], both as part of the national immunisation program (NIP).
The aim of this study was to describe and discuss the results of both, the BMSS and the massive vaccination strategies, on the epidemiology of BM in Cuban children and adolescents.
Methods
1023 cases of BM in the Cuban population between 1–18 years old reported nationwide by BMSS from January 1st 1998 to December 31st 2003 were included in the study, considering the date of the symptoms onset.
Children under 1 years old were not included in the study based on the criteria that they should be considered a special group for the following reasons: breast-feeding, immune maturation, they usually do not attend day care and do not have much community contact.
We defined a case of BM as "a clinical meningeal syndrome, through the identification of the causal agent directly by culture blood, petecchias, cerebrospinal fluid (CSF) analysis or indirectly by polymerase chain reaction, latex test or another rapid diagnostic test. The cases with negative culture or bacteriological test, but the cyto-chemical exam of the CSF suggested bacterial infection, were considered BM of "unknown aetiology" [11].
A total of 4 age groups which coincided with the main level of education in Cuba were selected: the group 1–5 year olds which attended day care centres (DCC) was considered as small children; 6–11 year olds were considered older children (Primary School). More than 11 year olds were considered adolescents, attending High School those 12–14 year olds and to College, Technical or Professional Education those 15–18 year olds.
The incidence rate (new cases per 100 000 population) and CFR (percent) were calculated according to the age group and to the main causal agent, using the data of estimated Cuban population of the Office for National Statistics in Cuba. The seasonality by monthly average for each causal agent and the frequency of cases who slept in overcrowded dormitories by age group were also estimated. Comparisons were carried out using Chi-squared tests with a significance level of 0.05.
Relative Risk (RR) was estimated and its confidence interval with p < 0.05 considered as significant, using either a Chi-squared Test or Fisher's Exact Test as appropriate. We considered exposed to community settings the cases attending DCC or boarding school, and not exposed those who do not attend these institutions. The national matriculation and registration of school children and adolescents during each school year were obtained from the Ministry of Education (MINED). RR in primary school children was not estimated because the boarding student matriculate between 1998–2003 was not available.
We used Epi-Info (version 6.2a) and Excel (version 5.1) for the statistical analysis, and Microsoft Word 2000 as the text processor program.
Results
A total of 1023 cases of BM between 1998–2003 were reported. The overall incidence rose significantly from 6.5 per 100 000 population in 1998 to 8.5 per 100 000 population in 1999, but it subsequently decreased to 4.2 per 100 000 population in 2003 (Table 1).
Table 1 Overall Incidence and Case-Fatality Rate (CFR) for bacterial meningitis* by age group in Cuba from 1998–2003.
Age group 1998 1999 2000 2001 2002 2003
1–5 yrs 16.8 15.3 11.1 8.5 4.7 6.2
6–11 yrs 3.3 7.4 4.1 3.7 3.3 4.6
12–14 yrs 1.6 5.9 3.7 3.5 4.3 4.1
15–18 yrs 3.7 4.5 4.5 2.9 1.7 2.0
Overall Incidence 6.5 8.5 5.8 4.6 3.4 4.2
Overall CFR 7.8 9.0 13.6 13.6 8.7 10.9
* Bacterial meningitis caused by Streptococcus pneumoniae, Neisseria meningitidis, Haemophilus influenzae type b, other identified bacteria and bacterial meningitis of unknown aetiology.
The incidence of all age groups increased significantly in 1999, with the exception of small children, where there was a peak in 1998 (16.8/100 000 population), with a subsequent decrease. The group 1–5 year olds showed the highest incidence of all. (Table 1)
An overall and by age group significant predominance of male (63.2%) over female (46.8%) was observed in all the groups (data not shown).
Spn, Hib and N. meningitidis were the main causal pathogens identified.
In 1998 Hib showed the highest incidence (3.0 per 100 000 population), but dropped significantly by approximately half (1.6 per 100 000 population) after the vaccination programme in 1999 with a conjugated vaccine (Vaxem-Hib®). The greatest decrease of incidence was observed in small children, from 10.7 per 100 000 population in 1999 to 5.7 per 100 000 population in 2000 (p < 0.05). Only 8 cases have been reported since 2002 (Table 2).
Table 2 Overall and by age group Incidence and Case-Fatality Rate (CFR) of bacterial meningitis according to the aetiology. Cuba. 1998–2003.
Streptococcus pneumoniae
Age group 1998 1999 2000 2001 2002 2003
1–5 yrs 1.6 2.8 2.9 2.0 1.9 2.3
6–11 yrs 0.3 0.9 0.9 0.6 0.2 0.7
12–14 yrs 0.4 0.6 1.1 0.5 0.4 0.7
15–18 yrs 1.4 0.8 0.8 0.3 0.3 0
Overall Incidence 0.9 1.3 1.4 0.9 0.7 1.0
Overall CFR 20.0 13.1 28. 6 26.9 15.0 30.0
Neisseria meningitidis
Age group 1998 1999 2000 2001 2002 2003
1–5 yrs 1. 1 0.8 0.9 1.2 0.7 0.5
6–11 yrs 0.5 0.5 0.5 0.4 0.3 0.3
12–14 yrs 0.2 0.8 0.6 0.7 0.9 0.4
15–18 yrs 0.5 0.6 0.9 0.1 0.1 0.1
Overall Incidence 0.6 0.6 0.7 0.6 0.5 0.3
Overall CFR 0 10.5 4.7 0 7.1 0
H. influenzae type b
Age group 1998 1999 2000 2001 2002 2003
1–5 yrs 10.7 5.7 1.6 0.5 0.6 0
6–11 yrs 0.7 0.4 0.2 0 0 0.1
12–14 yrs 0 0.2 0 0 0 0.2
15–18 yrs 0 0 0 0 0 0
Overall Incidence 3.0 1.6 0.5 0.1 0.2 0.1
Overall CFR 8.0 10.6 14.3 75.0 0 0
Unknown aetiology
Age group 1998 1999 2000 2001 2002 2003
1–5 yrs 4.6 6.2 4.7 4.3 1.3 2.2
6–11 yrs 2.7 5.6 2.1 2.6 1.9 2.8
12–14 yrs 1.2 3.4 1.1 2.2 3.0 2.1
15–18 yrs 1.7 3.0 3.4 2.3 1.2 1.7
Overall Incidence 2.7 4.8 2.9 2.9 1.8 2.3
Overall CFR 7.7 2.9 5.9 7.1 9.6 7.8
The overall incidence of pneumococcal meningitis was nearly 1 per 100 000 population, but a significant increase occurred in 1999 and 2000 (1.3 and 1.4 per 100 000 population respectively) and subsequently decreasing. A similar pattern in the incidence was observed in all age groups with the exception of adolescents from 15–18 years old. The small children showed the highest incidence of all groups, mainly in 1998 (2.8 per 100 000 population) and in 1999 (2.9 per 100 000 population), decreasing in 2001 and 2002, with a discreet subsequent increase in 2003 (2.3 per 100 000 population). Since the year 2000 Spn has shown the highest incidence of all the identified bacteria (Table 2).
Neisseria meningitidis was one of the three main causal pathogens of BM, but the overall incidence remained below 0.8 per 100 000 population, decreasing significantly towards 2003. All the age groups showed subtle variations of the incidence throughout the study period, however only small children exceeded 1 per 100 000 population (Table 2). The overall ratio of meningitis/septicaemia for all groups was 3 per 1 (Data not shown). Neisseria meningitis serogroup B was the only one isolated throughout the study in the Neisseria Laboratory at the Institute "Pedro Kourí" (IPK).
Regarding other identified bacteria, Staphylococcus sp., Streptococci (different from S. pneumoniae), E. coli and Salmonella were the most common pathogens. The frequency of this group of bacteria was 43% in small children, 28% in the older children, 13% in adolescents from 12–14 year olds and 16% in adolescents from 15–18 year olds. Gram-positive bacteria (63%) was significantly more frequent than the gram-negative (37%) (Data not shown).
The average CFR of BM was nearly 10.5%. Overall CFR showed variations, increasing significantly in 2000 (13.6%) and in 2001 (13.6%) when compared to 1998 (7.8%), with a subsequent significant decrease in 2003 (10.9%) (Table 1).
The most lethal pathogen was Spn with an average CFR = 27% (13–30%), followed by Hib with CFR = 10.7% (8–75%) and Meningococcus B with CFR = 4% (4.7–10.5%). The highest CFR (75%) observed in the period was caused by Hib in 2001, but no cases have been reported subsequently (Table 2).
Bacterial meningitis of unknown aetiology accounted for 500 cases throughout the period. The overall incidence showed a significant increase when compared in 1998 (2.7 per 100 000 population) to 1999 (4.8 per 100 000 population), as well as by age group. Only among small children a significant decrease was observed, from 4.6 per 100 000 population in 1998 to 2.2 per 100 000 population in 2003 (Table 2).
No seasonality was observed in pneumococcal and Hib meningitis. Only meningococcal meningitis showed seasonal behaviour with the highest peaks in September, October and March (Figure 1).
Figure 1 Seasonality of meningococcal meningitis in children and adolescents. Cuba. 1998–2003.
Fifteen percent of overall BM cases slept in overcrowded dormitories. This condition was observed in 10.1% of children, and increased significantly in adolescents from 12–14 year olds and 15–18 year olds (30.6% and 24.6% respectively) (Data not shown).
No association of BM with attendance to DCC or boarding school (including college, technical or professional education) was observed during the study. Only in 1998 and 1999 did the attendance to boarding high school show a strong association with disease (RR = 2.1; p > 0.05) (Table 3).
Table 3 Associations of bacterial meningitis with attendance to day care centres or boarding school. Cuba. 1998–2003.
Year Day Care Centres Boarding High School Boarding College, Technical or Professional Education
RR CI RR CI RR CI
1998 0.7 0.45–1.14 2.1 0.64–7.42 1.0 0.33–2.95
1999 1.3 0.84–2.02 2.1 1.02–4.50 1.0 0.42–2.68
2000 1.0 0.62–1.80 0.5 0.20–2.31 0.6 0.26–1.43
2001 0.4* 0.17–0.89 1.6 0.62–4.28 0.1* 0.03–0.68
2002 1.2 0.54–2.62 1.7 0.69–4.17 0.2* 0.04–1.03
2003 0.7 0.32–1.62 0.6 0.18–2.07 0.4 0.08–1.76
*p < 0.05 (chi-squared test or Fisher exact test as appropriate)
RR: Relative Risk
CI: Confidence Interval
Discussion
Prevention and control of infectious diseases, especially among children and adolescents, has been a big concern for the Public Health Ministry (PHM) in Cuba. Hitherto vaccines are the most promising prospect for the prevention of community acquired BM. Therefore widespread use of effective vaccines may induce a major impact in preventing the disease, especially when they are used during a long period of time as a part of the NIP, as occurred in Cuba, where a decreasing trend of BM has been documented after vaccination against meningococcus serogroup B and C, and more recently, Hib [4,9,10,12].
Cuban strategy to prevent these infections has been an initial vaccination campaign continued through the NIP.
The first intervention was carried out in 1979 with an A+C vaccine (Pasteur-Merieux) during the meningococcal disease epidemic caused mainly by serogroup C (50,0%) and B (34,3%). The target population was 3, 245 046 individuals from 3 months of age to 19 years old, considered at high risk of disease, achieving 78.2% vaccination coverage. This intervention decreased significantly the serogroup C (7.2%) in 1980, but was ineffective against serogroup B, therefore the incidence continued to rise, but having serogroup B as a predominant causal agent (78.4%) [4,12].
By the end of 1988, after having being evaluated in a double-blind, placebo-controlled trial, the Cuban anti-meningococcal BC vaccine (VA-MENGOC B-C®) was first administered to high risk age groups and subsequently to all the population from 3 months of age to 24 years of age (an estimated 2 million people or more) [4,9]. Since 1991 this vaccine has been included in the NIP for children of 3 months of life (2 doses, 2 months apart), decreasing the nation-wide incidence of meningococcal meningitis to below 1 per 100 000 population. For that reason, the contribution of meningococcal meningitis to the overall incidence of BM in Cuba is very small at present [13].
During this period, serogroup B Neisseria meningitidis (classified by a commercial sera Wellcome Diagnostic, England) has been the only one isolated bacteria and subsequently confirmed by the Neisseria Laboratory at the IPK, where the strain B4. P1. 19.15 has been the most frequently identified, using ELISA [14]. It is likely due to the fact that VA-MEMGOC B-C prevents the disease, but does not eliminate the nasopharyngeal carriage [9].
Shifts in the frequency and circulation of each bacteria and/or strain into a region, country or continent depends on the interaction of both the pathogen and the host within settled environmental conditions, nevertheless important changes may be induced by prolonged and massive vaccination programmes.
Small children are widely considered a high-risk group for BM, as it was confirmed in our study, where the highest incidence was observed in less than 6 year olds. [15,16]. In Cuba Hib and Spn were the leading pathogens of BM in this group since 1993, as a result of previous nationwide vaccination against meningococcus [5,12].
Due to the increase of infant mortality caused by Hib infections, the PHM decided to start using an available conjugate vaccine against Hib in 1999. The vaccine was initially used in a campaign for small children and continued through the NIP [10]. The incidence of Hib meningitis dropped by 47% after initial vaccination, especially in small children, coinciding with reports in other countries where those vaccines had also been used. Another beneficial effect of conjugated Hib vaccine is the elimination of carriage among vaccinated populations [17-19].
For these reasons, we considered that the vaccination against Hib contributed decisively to the overall decrease of BM in small children described above, and undoubtedly the development of such vaccines has been one of the most important events in the history of the prevention and control of infectious diseases in paediatrics [1,20].
Spn is not a common cause of meningitis epidemics and outbreaks [21]. We considered that the steady low incidence of meningococcal meningitis as well as the drastic and significant reduction of Hib meningitis in small children achieved during 1999 in Cuba, may have contributed to an increase of pneumococcal meningitis in 1999 and 2000, especially in children under 6 years old. We hypothesised that Spn might have filled somehow the ecological niche laid down abruptly by Hib after the incidence decrease subsequent to vaccination. Since then Spn is the leading cause of BM in Cuba as also occurs in many regions of the world [22].
The improvement of the BM surveillance system in Cuba increases substantially quantitative and qualitative information about this group of infections [12]. Despite these achievements, an important number of cases still remains of unknown aetiology due to previous antibiotic therapy.
Regarding the gender, we observed in our study that infections in male predominated those in females, contrarily to other author that point out non significant differences between males and females [1].
CFR depends on, among other factors, the host response, the virulence of the pathogen and the quality and timeliness of medical attention. Average CFR was nearly 10% and similar figures are reported in many developed countries, where it could be as low as 2% and as high as 25% [23,24], though in developing countries it may be higher. Pneumococcus was the most lethal germ, with CFR nearly 25%, and similar figures are reported in other regions of the world [16,25,26].
Two well-defined seasons are recognised in Cuba: the rainy (April to September) and the dry (October to March). The monthly average case distribution of Hib and pneumococcal meningitis, did not show seasonality differences, but meningococcal meningitis did, increasing the number of cases mainly in September, October and March, coinciding with reports of the literature [3]. In Cuba, September and October is the transitional period of the rainy season to the dry, but also it is the end of the summer vacations and the start of the school year, when many children and adolescents share the same environment (DCC and schools) and are in close contact with one another. Coincidence of these climatic, environmental and socio-cultural conditions, undoubtedly may contribute to the seasonal increase of meningococcal meningitis in our country.
Airborne transmission of infections can be favoured by some environmental conditions such as attendance at boarding institutions, overcrowding areas, poor ventilation and intimate contact [3,15,27].
In our study we observed that the overall percentage of cases sleeping in overcrowded dormitories was 15%, but in adolescents reached nearly 30%. It should be considered as an important risk for the transmission of BM in this group, and must call the attention of both, MINED and PHM, for the avoidance of these conditions, where possible, especially in DCC and boarding school.
On the other hand, as a nation-wide improvement of the Cuban educational system during the seventies, a huge number of boarding school (Colleges, Technical and Professional Education) were built nationwide. In these institutions the intimate contact and promiscuity among boarding students may increase airborne infections [3,15,27-29]. Nevertheless no strong association of BM with attendance to boarding school or DCC was observed, reinforcing the criteria of the beneficial effect of massive and systematic vaccination against two of the main BM causal pathogens.
Conclusion
Surveillance is a key subject for the control and prevention of infectious diseases and provides information about main epidemiological and microbiological aspects to put in place correct strategies.
Massive vaccinations against meningococcus, and Hib have decreased overall BM incidence in Cuba, and showing Pneumococcus as the leading causal pathogen and the most lethal.
The disease was more frequently observed among male children under 6 years of age.
Seasonality was only observed in meningococcal meningitis. Dry season and the start of the school year may also be contributing environmental conditions.
There was a weak association of BM with attendance to DCC and boarding school, and may be the effect of massive and systematic vaccination, which in our opinion are the most adequate and effective measures for the prevention and control of these infections.
Further massive vaccination initiatives with available anti-pneumococcal conjugated vaccine (when economically possible in Cuba) may decrease the incidence of BM even more. Surveillance systems will contribute towards the detection of the emergence of other causal pathogens or any other changes in the epidemiology of these infections.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
FDM conceived the study, participated in the design, performed the statistical analysis, and the draft of the manuscript and was the main responsible author.
APR participated in the design and draft of the manuscript. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
We are grateful to all colleagues at provincial level of the Bacterial Meningitis Surveillance System for their interest and effort in collecting and providing important information. Also we would like to thank Professor Diego Meneses, Dr. Angela Gala, Dr. Rafael Llanes, Dr. Solene Aoutin, Lic. Gabriel Díaz and Lic. Orquidea Biart for their helpful comments on an earlier version of this article.
==== Refs
Bonthius DJ Bahri K Meningitis and encephalitis in children. An update Neurol Clin N Am 2002 20 1013 1038 10.1016/S0733-8619(02)00016-6
Grimwood K Legacy of bacterial meningitis in infancy BMJ 2001 323 523 524 11546680 10.1136/bmj.323.7312.523
Organisation Mondiale de la Santé. Lutte contre les épidémies de méningite á méningocoque Guide pratique OMS 2 WHO/EMC/BAC/98.3.1999
Sierra G Campa C Valcárcel M García L Izquierdo L Sotolongo F Casanueva GV Rico CO Rodríguez CR Ferry MH Vaccine against group B Neisseria meningitidis: protection trial and mass vaccination. Results in Cuba NIPH Annals 1991 14 195 210 1812432
Dickinson F Pérez A Las meningoencefalitis bacterianas en la población infantil cubana:1998–2000 Rev Cubana Pediatr 2002 75 106 14
Bisgard KM Kao A Leake J Sterbel PM Perkins BA Wharton M Haemophilus influenzae invasive disease in the United States, 1994–1995: Near disappearance of a vaccine-preventable childhood disease Em Infect Dis 1998 4 229 238
John TJ Samuel R Balraj V John R Disease surveillance at district level: a model for developing countries Lancet 1998 352 58 61 9800768 10.1016/S0140-6736(97)10494-9
Thomas DR Salmon RL Westmoreland D Palmer SR Surveillance of influenza in Wales: interpreting sentinel general practice rates using contemporaneous laboratory data. Opportunities and limitations J Epidemiol Comm Health 1998 52 28 31
Pérez AE Dickinson FO Baly A Martínez R The epidemiological impact of anti-meningococcal B vaccination in Cuba Mem Inst Oswaldo Cruz 1999 94 433 40 10445998
Dickinson F Pérez A Galindo M Quintana I Impacto de la vacunación contra Haemophilus influenzae tipo b en Cuba Rev Panam Salud Publica/Pan Am J Public Health 2001 10 169 73
Quintana I Sotolongo F Llop A Cuevas I Martínez N Velázquez JC Alvarez R Martínez E Tamargo I Pérez A Martínez I Leiva T Programa Nacional de Prevención y Control de los Síndromes Neurológicos Infecciosos 1999 Havana: ServImpress
Pérez A Dickinson F Tamargo I Sosa J Quintana I Ortíz P Morales M Fumero A García M Quintana J Venero S Luján M Cabrera MV Font M Rodríguez O García R Rodríguez L Escobar ME Castro A Toledano E Urgellés G Rodríguez ME Marín A Llop A Batlle MC Fuentes K Pérez M Llanes R Guzmán D Gutiérrez O Molina I Toledo I Resultados y experiencias de la vigilancia nacional de meningitis bacterianas en Cuba Biotecnología Aplicada 2003 20 188 22
Pérez A Dickinson F Resultados del Programa Nacional de Inmunización Antimeningocócica BC en menores de 1 año en Cuba Rev Cubana Pediatr 1998 70 133 40
Sosa J Llanes R Guzmán D Quintana I Flores M Gutiérrez O Typing and susceptibility to Penicillin of Neisseria meningitidis isolated from patients in Cuba (1993–1999) Mem Inst Oswaldo Cruz 2001 96 523 25 11391426 10.1590/S0074-02762001000400014
Tan T Prevention of pneumococcal meningitis Current Infectious Disease Reports 2002 4 317 23 12126608
Solórzano F Miranda MG Díaz RD Meningoencefalitis bacteriana Enf Infec y Micro 2002 22 2 13
Adams WG Deaver KA Cochi SL Plikaytis BE Bell ER Broome CV Wenger JD Decline of childhood Haemophilus influenzae type b (Hib) in the Hib vaccine era JAMA 1993 269 221 6 8417239 10.1001/jama.269.2.221
Agudelo CI Muñoz N Hoz F de la y Laboratorios de Salud Pública Evaluación rápida del impacto de la vacuna contra Haemophilus influenzae serotipo b en Colombia Rev Panam Salud Publica/Pan Am J Public Health 2000 8 181 84
Ruocco G Curto S Savio M Laurani H Froch R Vacunación contra Haemophilus influenzae tipo b en el Uruguay: experiencia e impacto Pan Am J Public Health 1999 5 197 99
Schuchat A Robinson K Wenger JD Harrison LH Farley M Reingold AL Lefkowitz L Perkins BA for the Active Surveillance Team Bacterial meningitis in the United States in 1995. Active Surveillance Team N Engl J Med 1997 337 970 6 9395430 10.1056/NEJM199710023371404
Peltola H Prophylaxis of Bacterial Meningitis Infectious Disease Clinics of North America 1999 13 685 710 10470562 10.1016/S0891-5520(05)70101-X
Klein JO Quie PG, Kass EH The epidemiology of pneumococcal disease in infants and children The pneumococcus and the pneumococcal vaccine 1982 Chicago: Chicago Press 64 71
Sáez-Llorens X Mc Cracken GH Bacterial meningitis in children Lancet 2003 361 2139 48 12826449 10.1016/S0140-6736(03)13693-8
Razonable RR Keating MR Meningitis [Monography in Internet]. Emedicine. 2002 Sept [Cited September 6th 2004]
Scott G The prevention of pneumococcal disease in children N Engl J Med 2001 345 1177 83 11642234 10.1056/NEJM200107263450420
Ortquist A Pneumococcal disease in Sweden: experiences and current situation Am J Med 1999 107 44 9 10.1016/S0002-9343(99)00101-1
Sáez-Llorens X McCracken GH Jr Bacterial meningitis in neonates and children Infect Dis Clin North Am 1990 4 623 44 2277192
Baltimore RS Jenson HB Meningococcal vaccine: new recommendations for immunization of college freshmen Curr Opin Pediatr 2001 13 47 50 11176243 10.1097/00008480-200102000-00008
Shapiro ED Prophylaxis for contacts of patients with meningococcal or Haemophilus influenzae type b disease Pediatr Infect Dis 1982 1 132 38 6757888
|
16288649
|
PMC1299326
|
CC BY
|
2021-01-04 16:28:16
|
no
|
BMC Infect Dis. 2005 Nov 15; 5:103
|
utf-8
|
BMC Infect Dis
| 2,005 |
10.1186/1471-2334-5-103
|
oa_comm
|
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-981626643610.1186/1471-2334-5-98Research ArticleCerebrospinal fluid HIV infection and pleocytosis: Relation to systemic infection and antiretroviral treatment Spudich Serena S [email protected] Annelie C [email protected] Nicole D [email protected] Teri J [email protected] Christos J [email protected] Steven G [email protected] Ellen E [email protected] Richard W [email protected] Department of Neurology, University of California San Francisco, USA2 Department of Medicine, University of California San Francisco, USA3 Gladstone Institute of Virology and Immunology, San Francisco, USA4 ViroLogic, Inc., South San Francisco, USA2005 2 11 2005 5 98 98 27 5 2005 2 11 2005 Copyright © 2005 Spudich et al; licensee BioMed Central Ltd.2005Spudich et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Central nervous system (CNS) exposure to HIV is a universal facet of systemic infection. Because of its proximity to and shared barriers with the brain, cerebrospinal fluid (CSF) provides a useful window into and model of human CNS HIV infection.
Methods
Prospective study of the relationships of CSF to plasma HIV RNA, and the effects of: 1) progression of systemic infection, 2) CSF white blood cell (WBC) count, 3) antiretroviral therapy (ART), and 4) neurological performance. One hundred HIV-infected subjects were cross-sectionally studied, and 28 were followed longitudinally after initiating or changing ART.
Results
In cross-sectional analysis, HIV RNA levels were lower in CSF than plasma (median difference 1.30 log10 copies/mL). CSF HIV viral loads (VLs) correlated strongly with plasma VLs and CSF WBC counts. Higher CSF WBC counts associated with smaller differences between plasma and CSF HIV VL. CSF VL did not correlate with blood CD4 count, but CD4 counts <50 cells/μL associated with a low prevalence of CSF pleocytosis and large differences between plasma and CSF VL. CSF HIV RNA correlated neither with the severity of the AIDS dementia complex (ADC) nor abnormal quantitative neurological performance, although these measures were associated with depression of CD4 counts.
In subjects starting ART, those with lower CD4 counts had slower initial viral decay in CSF than in plasma. In all subjects, including five with persistent plasma viremia and four with new-onset ADC, CSF HIV eventually approached or reached the limit of viral detection and CSF pleocytosis resolved.
Conclusion
CSF HIV infection is common across the spectrum of infection and is directly related to CSF pleocytosis, though whether the latter is a response to or a contributing cause of CSF infection remains uncertain. Slowing in the rate of CSF response to ART compared to plasma as CD4 counts decline indicates a changing character of CSF infection with systemic immunological progression. Longer-term responses indicate that CSF infection generally responds well to ART, even in the face of systemic virological failure due to drug resistance. We present simple models to explain the differing relationships of CSF to plasma HIV in these settings.
==== Body
Background
Frequent abnormalities in cerebrospinal fluid (CSF), including increased white blood cells (WBCs), were recognized early in the AIDS epidemic, not only in patients examined toward the end of their course who suffered neurological complications [1], but also in those systemically and neurologically asymptomatic [2,3]. Indeed, these observations were among the first indicators that the central nervous system (CNS) is an early and common target of systemic HIV infection. While initial studies applying quantitative HIV RNA measurements to the CSF suggested correlation between the CSF HIV RNA (viral load, VL) and the AIDS dementia complex (ADC) [4], subsequent reports have shown that HIV can be found in the CSF throughout the course of infection, beginning with primary infection [5], and that other factors, including the progression of immune dysfunction, are likely important in the development of ADC [6-8]. This has raised the fundamental question of why HIV causes brain dysfunction, manifesting as ADC, only late in the course of infection and only in some individuals [9]. Additionally, because the brain and CSF are separated from the blood by barriers to the transfer of virus, immune defenses and antiviral drugs, there has been considerable concern as to whether local infection in these 'compartments' might be isolated from host defenses and antiviral therapy (ART), leading to both viral persistence and local selection of resistance [10-12]. While CSF and brain infections by HIV are not identical, examination of this easily sampled fluid provides a window into CNS infection [13].
In order to better interpret CSF findings, it is essential to understand what factors contribute to elevated CSF HIV RNA concentrations. How do systemic infection and its progressive damage to the immune system affect the VL in a non-lymphatic compartment like the CSF space? How does CSF infection respond to ART? What is the origin and importance of the CSF cell reaction detected as CSF lymphocytic pleocytosis and how does this cell reaction respond to ART? What is the effect of ART on neurological function in subjects presenting with ADC?
To address these questions and better understand the relationship of the CSF HIV RNA levels to other aspects of infection and clinical findings, we undertook a prospective study of CSF in a broad range of HIV-infected subjects using two complimentary approaches. The first involved cross-sectional analysis of a clinically diverse subject sample. The second longitudinally followed subjects initiating ART. In addition to its direct clinical implications, this longitudinal approach used treatment as an 'experimental' intervention to dissect dynamic aspects of the relationships among these study variables.
Methods
Subjects and Protocols
One-hundred subjects were entered into these studies between November, 1996 and June, 2001 in the context of protocols approved by the University of California, San Francisco (UCSF) Committee on Human Research (CHR); follow-up on a few continued until December 2004. Informed consent was obtained from all subjects. In the case of one subjects with ADC, consent was also obtained from his sister with durable power of attorney. Subjects were excluded if they suffered HIV-related or other active CNS diseases except ADC. The cross-sectional analysis targeted a total of 100 subjects, including the baseline observations of subjects starting treatment, along with a previously-reported group who stopped treatment [14,15]. Subjects were clinically stable with the exception of six presenting with a new diagnosis or progression of ADC at the time of study entry.
Treatment decisions were independent of this CSF study. New therapies (either starting de novo or representing a change of drugs) were prescribed by subjects' primary care-giver or determined by another clinical trial. Exceptions were five subjects who entered an open-label study of high-dose abacavir (600 mg twice daily) which was completed before abacavir was licensed for clinical use [16,17]. The first subject in this small protocol substituted abacavir for one of the drugs in his existing regimen, while the remaining subjects began this nucleoside reverse transcriptase inhibitor (nRTI) in the context of novel combination ART. Treatment protocols sought to enter subjects with and without ADC, and with plasma VLs ≥ 50,000 HIV RNA copies/mL (cpm), although exceptions were made to this virological entry criterion. Baseline lumbar punctures (LPs) were performed within one week before initiating ART and carried forward as time zero. Most subjects underwent their initial 4–5 LP on-drug studies during the first month, with later assessments at approximately 3 and 6–12 months and then one to three times yearly thereafter.
LP and CSF analysis
CSF was obtained for study purposes rather than for clinical diagnosis and was processed in standardized fashion as previously described [18,14]. CSF and plasma from each visit were analyzed for HIV RNA concurrently. At the time of the first LP, CSF was also analyzed for neurosyphilis (VDRL) and cryptococcal antigen (all negative). CSF cell counts, differential, protein and albumin levels along with measurement of blood albumin and CD4+ and CD8+ T lymphocyte counts by flow cytometry were performed using routine clinical methods in the San Francisco General Clinical Laboratory.
Clinical evaluation
Each subject had a baseline clinical evaluation, which included a general and neurological assessment and medical history review; none had current or prior cryptococcal meningitis or any other opportunistic neurological disorder. If indicated by neurological symptoms or signs, subjects had neuroimaging studies to assess for confounding conditions. Subjects underwent a standardized, ADC-focused neurological evaluation leading to ADC diagnosis and staging [19-22]. Diagnosis of ADC conformed to criteria for the AIDS-related cognitive/motor complex outlined by the American Academy of Neurology Task Force [23]. In the presence of any static neurological condition that might interfere with designation of AIDS related cognitive or motor dysfunction (for example, prior head trauma or psychiatric diagnosis), no ADC scale was assigned. Subjects underwent brief quantitative performance testing with a battery of four tasks (timed gait, finger tapping with the dominant hand, grooved pegboard placement with the non-dominant hand, and Digit Symbol test from the WAIS-R) yielding a combined normalized score derived from the mean of individual Z-scores, the quantitative neurological performance Z-score on four tests (QNPZ-4 score), as previously described [24,20,22].
Virological methods
HIV RNA was measured in cell-free CSF and plasma by the Roche Amplicor Monitor assay (versions 1.0 and 1.5, Roche Diagnostic Systems, Inc., Branchburg, N.J) using the standard and Ultrasensitive extraction methods. The latter has a quantitation limit of 50 and a detection limit of approximately 20 HIV RNA cpm. We used results in the range of 20 – 50 cpm for data reporting and analysis, and assigned a default 'floor' value of 19 (log10 1.28) cpm for values below the detection limit. Concurrent paired CSF and plasma samples were treated identically and run at the same time. HIV RNA concentrations were transformed to log10 values for all analysis. Because limited serial sampling during the acute phase of treatment-induced viral decay precluded complex modeling, estimates of acute-phase HIV-RNA half-lives were derived by assuming simple exponential decay during an initial phase (days 0 – 11) as previously described [25]. While the acute phase of plasma viral decay is shorter than 11 days, we used this extended period for rough comparison because of the limited sampling in some subjects. The acute decay rate, λ, for each subject was estimated by least-square regression of measurement time (including baseline) on a subject's log10 HIV-RNA values. The subject's acute-phase HIV-RNA half-life was log102/λ.
Antiretroviral drug resistance
Antiretroviral drug resistance was assessed by both phenotypic and genotypic methods in selected subjects using the PhenoSense™ HIV assay (ViroLogic Inc., South San Francisco, CA) to analyze functional susceptibility in a recombinant assay and mutations in the reverse transcriptase (RT) and protease (PR) regions [26]. Phenotypic susceptibility results were reported as fold-change in the 50% inhibition concentration (IC50) relative to a wild-type virus reference standard, while genotypic analysis compared RT and PR sequences in the blood or CSF HIV populations to the reference wild-type (e.g., NL4-3).
Statistical analysis
Because of the skewed distributions of several study variables, unless otherwise indicated, the median and interquartile range (IQR; 25th percentile – 75th percentile) were used for descriptive statistics, and nonparametric tests were used for comparisons. In the cross-sectional analysis, all p-values were two-sided, with p-values < 0.01 considered significant. Statistics were performed using SPSS 11.5 (SPSS Inc., Chicago, IL), or Prism 4.0 (GraphPad Software Inc, San Diego, CA).
Results
Study subject demographics
Of the 100 subjects in the cross-sectional study, 65 were recruited into a cross-sectional only group and studied only once, 26 were in a longitudinal treatment group and were followed with serial LPs, and nine subjects were in a longitudinal Structured Treatment Interruption (STI) group and are reported elsewhere [14,15]. However, as two of these nine STI group subjects were studied after they restarted treatment, they were also included among the longitudinal treatment group, bringing the total number of patients in the longitudinal treatment group to 28. Table 1 provides a summary of the salient clinical and laboratory variables in the 100 subjects and also divides them according to ART treatment status. Additionally, Table 1 summarizes data from a subgroup of patients on ART who were considered treatment failures, as defined by plasma HIV RNA concentration >500 RNA cpm.
Table 1 Baseline Characteristics of Study Subjects, including subgroups.
Sex: HIV-1 RNA
N Age M:F CD4+ Duration of Infectiona CDC Stage C3b ADC Stage QNPZ-4 Score Plasma CSF P-C log10 Diff. CSF WBC
(yrs) (ratio) (cells/mm3) (mean yrs +/- SD) (percent) (log10RNA copies/mL) (cells/mm3)
Total 100 39.0 (36.0–45.0) 91:9 181.5 (48.8 – 285.3) 9.7 (+/- 5.60) 70.3% 0 (0 – 1) -0.50 (-1.53 – 0.21) 4.73 (3.52 – 5.15) 2.74 (1.48 – 4.00) 1.30 (0.19 – 2.32) 1.0 (0.0 – 4.0)
ADC ≥ 1 = 30.0%
Subdivision by Treatment Group and Effect
Off Treatment
46 38.0 (33.5 – 43) 42:4 195.0 (33.5 – 307.5) 7.5 (+/- 5.80) 65.2% 0 (0 – 0.5) -0.46 (-1.35 – 0.26) 4.93 (4.53 – 5.49) 3.61 (2.57 – 4.40) 1.19 (0.42 – 21.7) 2.0 (0 – 11.3)
ADC ≥ 1 = 23.1%
On Treatment (Total)
54 43.0 (38.0-29.3) 49:5 181.5 (76.0 – 275.5) 11.09 (+/- 5.08) 75.9% 0.5 (0 – 1.0) -0.51 (-2.62 – 0.20) 3.40 (1.85 – 4.80) 1.66 (1.28 – 3.07) 1.37 (0.0 – 2.42) 1.0 (0.0 – 2.0)
ADC ≥ 1 = 36.6%
On Treatment Failures
36 40.0 (38.0 – 47.8) 32:4 166.5 (48.3 – 276.5) 11.24 (+/- 4.18) 75.0% 0.5 (0 – 1.0) -0.42 (-2.40 – 0.25) 4.73 (3.99 – 5.04) 2.45 (1.45 – 3.64) 2.19 (1.29 – 2.59) 0.5 (0.0 – 2.8)
ADC ≥ 1 = 37.0%
Values are medians with IQR in parentheses beneath, unless noted. P-Clog10 Diff. is the difference in plasma and CSF log10 HIV concentrations.
aData available for 49 of the 100 subjects.
bData available for 81 of the 100 subjects; categorization based on the 1993 Centers for Disease Control HIV classification system.
Reflecting the demography of the local epidemic, over 90 percent of our subjects were men, with a median age of 39 years. The majority of the cross-sectional only sample was on ART, though VL was undetectable in plasma in only eight of these 65 subjects, reflecting a bias toward entry of subjects with virological failure.
While most of the 28 subjects in the longitudinal treatment group were ART-naive or had limited prior therapy, seven were on therapy, changing or adding one or more drugs at baseline. Twenty-four of 80 subjects without confounding neurological conditions (30%) were diagnosed with ADC: 13 subjects with ADC Stage 1, 8 subjects with ADC Stage 2, and 3 subjects with ADC Stage 3. Most had stable neurological impairment and clinically 'inactive' CNS disease (exceptions are discussed below). The prevalence of ADC was the principal reason for the overall median QNPZ-4 score below "normal".
Cross-sectional analysis
CSF HIV RNA
In the cross-sectional evaluation, HIV RNA was characteristically lower in CSF than in plasma, and the VL differences between the two compartments, which we express here and below as the ΔPlasma:CSF (log10 plasma HIV RNA – log10 CSF HIV RNA), varied widely, ranging from -1.32 to 4.08 log10 cpm (Table 1). CSF VL was higher than plasma VL in only 11 of 100 subjects, with a difference of >0.50 log10 cpm found in only 3 of these 11 subjects. Figure 1 shows the distribution of the plasma (A) and CSF (B) HIV RNA in relation to the blood CD4 cell counts in the 100 subjects. Division of the results into CD4 quartiles shows that the cohort did not distribute evenly among CD4 values, so that one quarter of the subjects had CD4 counts below 49 cells/μL.
Figure 1 Baseline distributions of HIV concentrations and CSF WBCs in relation to blood CD4 cell counts in the 100 subjects analyzed cross-sectionally. Individual panels show: (A) plasma HIV RNA, (B) CSF HIV RNA, and (C) CSF WBC counts in relation to blood CD4 counts. The vertical dotted lines in each panel separate the subjects into blood CD4 quartiles. The Symbol key appears in the bottom right panel, red boxes indicate patients off treatment, while blue circles indicate those on treatment.
The treated subjects had lower VLs in both compartments (Table 1, Figure 1; p < 0.001 for both; Kruskal-Wallis test). To assess the effects of ART on CSF HIV in the setting of treatment failure, we analyzed the results in the subjects on ART who had plasma HIV VLs above 500 cpm and include this subgroup in Table 1. The median plasma VL in this subgroup was similar to that of the off treatment group, though these groups differed statistically (p = 0.017, Mann-Whitney). More notably, the median CSF values differed by more than 10-fold (p = 0.001), and therefore the median ΔPlasma:CSF of the failure group was greater than the median for subjects off treatment. This post hoc analysis raises the question of whether 'failed' treatment might alter the relationship of CSF to plasma HIV, and indicates that despite plasma HIV escape, treatment still had an effect on CSF.
Overall, CSF HIV RNA correlated with only two other variables examined, the plasma VL and the CSF WBC count (p < 0.001 for both, Spearman's rho 0.629 and 0.516, respectively). In contrast, CSF HIV did not correlate with the blood CD4+ cells, ADC stage or QNPZ-4 scores. However, as shown on Figure 1, plasma VLs were highest in subjects in the first CD4 quartile (<49 cells/μL). The plasma VL in this quartile was 5.03 log10 cpm with IQR 4.67 – 5.57, while the median plasma VL of the remaining 75 subjects was 4.35 log10 cpm with IQR 2.92 to 4.98 (p = 0.001). In contrast, the CSF VLs of subjects in this first quartile (median 2.75 log10 cpm, IQR 2.15 – 3.20) did not differ from those of the remaining 75 subjects (median 2.64 log10 cpm, IQR 1.28 – 4.19) (p = 0.650). The combined effect of higher plasma VL and similar CSF VL in the first quartile compared to the other subjects resulted in a greater median ΔPlasma:CSF for subjects in the first quartile (2.25 log10 cpm, with IQR 1.75 – 2.95), relative to the median for the remaining 75 subjects (0.800 log10 cpm, IQR of 0.00 – 1.97) (p = 0.001).
CSF WBC counts
Twenty-four subjects had abnormal CSF WBC counts (>5 cells/μL), composed of 85–100 percent lymphocytes, with the remainder mononuclear cells. All 24 were asymptomatic, despite a median count of 15 cells/μL (IQR 9 – 32 cells/μL, range 6 – 66 cells/μL). The CSF WBC count did not significantly correlate with plasma VL.
Figure 2 shows the relationship between CSF WBC count and plasma and CSF HIV VLs using a three-dimensional plot. The highest CSF VLs were in subjects with both pleocytosis and high plasma VLs (generally ≥ 4.0 log10 cpm). These were also the subjects with the highest CSF VLs. By contrast, many subjects with similarly high plasma VLs without pleocytosis had lower CSF VLs. Three subjects with CSF VLs that substantially exceeded those of plasma are indicated by subject number in Figure 2 and are discussed below. Subjects with CSF WBC counts ≥ 10 cells/μL had a median VL difference between the two fluids of 0.150 log10 cpm (IQR -0.115 – 0.425 copies) – far below the difference seen in the group overall.
Figure 2 Three-dimensional plot showing relationships among plasma and CSF HIV RNA concentrations and CSF WBC counts. Subjects with CSF pleocytosis had highest CSF VLs. In turn, higher CSF VLs were noted chiefly in those with elevated plasma HIV (>10,000 cpm). The numbers near three of the data points identify the three subjects with substantially higher CSF than plasma HIV RNA (see text).
CSF WBC counts showed only a modest correlation with CD4 counts (p = 0.01; rho 0.254), Figure 1C shows that this correlation related chiefly to differences between the subjects in the lowest CD4 quartile (less than 50 cells/μL) and the remaining subjects with higher CD4 counts. The median WBC count for first quartile was 0 cells/μL (IQR 0 – 1.0) while that for the remaining subjects was 2.0 cells (IQR 0 – 8.3 cells) (p = 0.009; t-test).
Neurological status
Neither the ADC stage nor the QNPZ-4 score correlated with the CSF or plasma VLs across the cross-sectional group, though these two measures correlated strongly with each other (p < 0.001, rho -0.746), and both also correlated with the CD4 count (p < 0.001 for both, rho = -0.515 for ADC stage and 0.467 for QNPZ-4). For all ADC subjects, the CD4 median was 67.5 cells/μL (IQR 28.0 – 124.5); for the untreated ADC subgroup, the median was 35 (IQR 14.5 – 67.5) cells/μL, and for those on ART, the median was 95 cells/μL (IQR 36 – 205).
Only 6 subjects were judged to have active ADC. Two of these were treatment failures with active disease despite ART, while the other four were off ART at presentation. Among the six were the three subjects designated as outliers in Figure 2 because of VLs higher in CSF than in plasma: subject 4034 with ADC Stage 1, and subjects 4032 and 4033 with Stage 2. While two had elevated CSF WBC counts consistent with the general positive correlation between of high CSF VL and CSF WBC count, subject 4032 did not. His CSF VL was high (19,700 cpm), his plasma HIV level was nearly tenfold lower (2,800 cpm), and the usual Δplasma:CSF was reversed despite acellular CSF. While he was prescribed nelfinavir, abacavir and zidovudine, it was suspected that he was not consistently taking full dosage, and genotypic resistance studies showed no evidence of significant resistance-associated mutations in the HIV PR or RT regions (not shown).
Longitudinal treatment studies
Longitudinal analysis involved 28 subjects who were followed with repeated LPs after initiating or modifying treatment. This is an extension of our earlier published study, adding 13 subjects and prolonging the period of follow-up for several of the 15 subjects previously described [25]. Twenty of these subjects were either treatment-naïve, had limited ART exposure, or had been fully suppressed in the past before stopping therapy and were therefore anticipated to respond well to ART. The remaining eight subjects had failed their previous treatment and were either changing therapies or restarting ART after a hiatus, with addition or substitution of one of more drugs at entry into the study.
Subjects underwent multiple LPs (median 6, IQR 5–8). Figure 3 shows the course of their HIV RNA levels in plasma (A) and CSF (B) along with changes in CSF WBCs (C) during the follow-up. At baseline, the plasma VLs were both higher and within a narrower range than those of CSF.
Figure 3 Responses to ART. The panels show the individual subject plots of changes in the plasma (A) and CSF (B) HIV RNA concentrations and (C) WBC counts after treatment with the time axis broken into three segments showing initial, intermediate and longer-term outcomes.
Early-phase CSF virological responses
The graphs in Figure 3 are divided into 3 temporal segments. Visual comparison of the initial segment (days 0–11) suggests that early viral decay was slower in CSF than in plasma for some subjects. Using previously described methods that apply linear regression to the log10 HIV RNA values in CSF and plasma [25], and restricting comparison to subjects who began multidrug regimens and exhibited rapid initial plasma decay, we derived CSF:plasma decay ratios in 18 of the subjects who had at least 3 LPs between days 0 and 11.
The CSF:plasma HIV decay ratios were variable (median 0.82, IQR 0.41–1.10). Exploration of the relationship of these ratios to baseline variables showed a high correlation with only the blood CD4 T lymphocyte count (p < 0.000, rho -0.744), and that neither the baseline VLs themselves, the CSF WBCs, nor the ADC stage had a significant effect. Figure 4 shows the relationship between the CSF:plasma viral decay ratio and the baseline CD4 counts and includes a regression line that shows equal decay in plasma and CSF (a CSF:plasma decay ratio value of 1) near CD4 = 250 cells/μL, with slower decay at lower CD4 counts. This decay difference related to CD4 count suggests a change in the character of CSF infection with more advanced systemic infection.
Figure 4 Relation of CSF:plasma early-phase decay ratio to baseline blood CD4 cell counts. The regression line and 95% confidence intervals were plotted after censoring one subject with CD4 = 1,140 cells/μL. The p-value and r2 of this regression analysis are shown on the figure, while the results of nonparametric analysis are discussed in the text. The horizontal broken line designates the point at which plasma and CSF decay are equal (ratio of 1) and the vertical broken line signals the point where this crosses the regression line – near a blood CD4 count of 250 cells/μL.
Longer-term virological and WBC responses in CSF
Despite slower initial decay in CSF compared to plasma in some subjects, the longer-term effects of ART on CSF HIV RNA in this group were excellent, and all subjects reached or approached the limit of detection over the period of observation (Figure 3A and 3B). As shown in Figure 3C, treatment also eliminated the CSF pleocytosis in all subjects with elevated baseline WBC counts.
Particularly notable in the treatment group were five subjects who achieved CSF HIV suppression despite persistent plasma viremia. Their CD4 counts were similar to the larger group (median 205 cells/μl, range 77 – 269). Four were neurologically normal, while the fifth (subject 5007) had a diagnosis of ADC based on longstanding and clinically static myelopathy. The plasma and CSF HIV and WBC changes for these five subjects with dissociated longer-term responses are shown in Figure 5 (upper panels), along with results of phenotypic resistance testing for the drugs that they were taking (lower panels). The antiretroviral medication histories and genotypic resistance mutations detected in the two fluids of these five subjects are presented in Table 2 [see Additional file 1]. All were treatment-experienced when entering the study; three (4001, 5001 and 5007) were on therapy and changing or modifying their regimens, while the remaining two (4015 and 4030) were off ART at the study start and were initiating new regimens.
Figure 5 Dissociated CSF and plasma HIV RNA responses in five subjects. Each of these subjects achieved near or full CSF viral suppression despite an incomplete plasma response. The top panels of each pair show CSF and blood HIV RNA and CSF WBC values (A). The lower panels graphically depict the phenotypic resistance profiles as fold change in susceptibility to the drugs these subjects were taking during the study compared to reference wild type on a log10 scale [26]. See the text for discussion.
The results of genotypic and phenotypic drug resistance analysis of plasma and CSF indicate that persistent plasma viremia was associated with drug resistance that was either demonstrable at baseline (subjects 4001, 5001 and 5007) or emerging during the period of observation (4015 and 4030). In two subjects (5001 and 5007) with baseline resistance, minor differences in genotypic mutations suggested at least partial compartmentalization, with CSF viruses more drug-susceptible than the predominant plasma quasispecies. Phenotypic drug susceptibility testing in Subject 5001 (Figure 5B) showed resistance to drugs in his regimen in both compartments, including resistance to abacavir. Minor differences in resistance mutations in the two fluids (Table 2 [see Additional file 1]) were likely insufficient to explain his greater CSF response. Genotypic analysis in Subject 5007 (Table 2 [see Additional file 1]) suggests a mixed population of drug susceptible and resistant viruses in CSF samples (for example M41M/L, L74L/V and T215T/N/S/Y) compared to a predominantly resistant population in plasma (for example M41L, L74V and T215Y). Perhaps greater susceptibility of a 'compartmentalized' CSF virus population contributed to the greater CSF response (Table 2 [see Additional file 1] and Figure 5C). Interestingly, the highly resistant plasma HIV population did not 'overflow' into the CSF and completely alter its resistance profile.
The other three (4001, 4015, and 4030) showed no evidence of compartmentalization at baseline, with similar susceptibility in both plasma and CSF virus populations. In Subject 4001 (Figure 5A), high-level resistance to lamivudine, saquinavir and ritonavir raises the question of whether the therapeutic effect on CSF may have related principally to stavudine, though this drug was insufficient to suppress plasma HIV. The initial samples of both plasma and CSF from Subject 4015 (Figure 5D) showed mixed populations of lamivudine resistance (M184M/V), from which emerged the resistant quasispecies (M184V) at the time of viral rebound (Table 2 [see Additional file 1]). In Subject 4030 (Figure 5E) phenotypic and genotypic resistance testing showed nearly identical susceptibility in CSF and plasma, and wild-type genotypes in both compartments. At day 85 and afterwards, his plasma HIV showed increasing resistance to lamivudine, nelfinavir and efavirenz (resistance to nevirapine increased similarly, not shown), with susceptibility only to stavudine, though genotyping showed the emergence of a stavudine mutation, T215Y. In addition to the disproportionate CSF HIV response, CSF WBC counts were also suppressed in these subjects. This was most remarkable in subjects 4015, with more than 45 cells/μL, and 5001, with 33 cells/μL at baseline; in both, the pleocytosis resolved despite sustained plasma HIV.
Neurological responses to ART
In the 24 treatment-group subjects without active ADC at baseline, QNPZ-4 scores were stable or showed small increases over the course of observation (not shown). The four subjects in the treatment group who presented with active ADC (5002, 4033, 4013, and 4034) showed both clinical improvement and distinct increases in this performance score in response to initiation or change of antiretroviral therapy. The clinical histories, antiretroviral therapy regimens, laboratory measurements, and QNPZ-4 scores of these subjects are presented in Figure 6. Of note, genotypic resistance testing, available at baseline for Subject 4034 (Figure 6D) who developed ADC on treatment with an unusual regimen, showed concordance in CSF and plasma with no resistance mutations in the RT and only L63P, A71A/V and V77I changes in the PR. These findings are consistent with insufficient drug potency and poor drug penetration into the CSF, and might suggest that the CNS served as a major site of viral replication in this subject. In response to a change to abacavir, nevirapine and indinavir/ritonavir, he achieved virological suppression in both and an improvement in the speed and clarity of his cognition.
Figure 6 Longitudinal follow-up of four subjects presenting with new-onset or progressing ADC. The upper panel of each pair shows the plasma and CSF VL responses along with CSF WBC changes, and the lower panels show treatment effects on the QNPZ-4 scores and the blood CD4 counts. The table below each graph indicates clinical features of each respective subject. Antiretroviral medications considered able to penetrate the CSF [10] are indicated in bold font. The key to the symbols for all graphs are shown to the left of panel set (A).
Discussion
The CSF found in the ventriculo-leptomeningeal space is separated from systemic sources of virus, immune defenses, and antiviral drugs, and is easily sampled by lumbar puncture. Thus, study of CSF HIV infection can serve a valuable role in understanding the dynamics and mechanisms of infection within an isolated tissue compartment. Further, CSF infection serves as both a model of and window into brain infection, providing important insight into viral neuropathogenesis. This CSF 'compartment' can be viewed as parallel to the brain compartment, sharing some of its barrier features. However, exchanges between CSF and blood may differ from those between brain and blood, and as a result, tissue responses may differ from CSF responses in important ways [27]. More directly, the CSF may reveal brain processes as a consequence of its intimate contiguity and function as an extracellular 'sink' where molecules produced in the brain (or along its perivascular spaces) that have diffused into the CSF can be sampled [8].
The studies described here used complimentary cross-sectional and longitudinal approaches to show that CSF infection is a nearly universal facet of the ecology of systemic HIV infection. CSF HIV RNA concentrations correlated with those of plasma, but were characteristically lower. The median difference between plasma and CSF HIV VLs (Δplasma:CSF) was 1.30 log10 cpm though this difference also varied considerably, ranging from -1.32 to 4.08 log10 cpm. Our results extend previous reports [4,6,7,28-31] and focus on the factors that modify CSF infection in relation to plasma VL. CSF infection is importantly influenced by the CSF WBC response, the degree of systemic immunological progression as measured by blood CD4+ T cells, antiretroviral treatment and drug resistance, and the presence of active ADC.
CSF WBCs
CSF WBC counts were commonly elevated in this series, confirming findings reported earlier in the epidemic (e.g., [2]). Our steady-state cross-sectional and treatment-related longitudinal observations confirm that this CSF pleocytosis is directly linked to HIV infection itself, rather than to another cause such as undiagnosed opportunistic infection. Specifically, the cross-sectional analysis showed that CSF WBC counts were highly correlated with CSF HIV RNA concentrations (Figure 2) though not with the plasma VL. A similar association of CSF HIV with CSF WBC count has previously been reported by several groups [32,8,33]. The longitudinal studies of treatment indicated this association even more clearly, revealing resolution of baseline pleocytosis in all those beginning ART (Figure 3C). This association is also supported by our previous report that about half of subjects undergoing STI develop a brisk CSF pleocytosis upon stopping therapy [14].
While this combined experience indicates that CSF VLs and WBC counts are related, it begs the fundamental mechanistic question of whether the CSF WBCs actually contribute to raising the CSF VL, or alternatively, represent only a response to high CSF virus. We discuss these two alternative hypotheses in relation to models of CSF infection below.
Change in the character of CSF infection with disease progression
Two of the observations reported here show a change in the relation of CSF infection to systemic infection as the latter progresses (as indicated by reduction in blood CD4 lymphocyte counts). First, in the cross-sectional study the Δplasma:CSF values in subjects with CD4 counts below 50 cells/μL (the first CD4 quartile of the group) were significantly larger than those of the remaining subjects with more preserved CD4 counts. This was related to higher plasma VLs, but without concomitant increase in CSF VLs; rather CSF VLs for these subjects were similar to those of the remaining subjects. This quartile also had lower CSF WBC counts. This observation confirms the impression of earlier investigators that CSF cell counts decrease in those with more advanced systemic infection [2]. We discuss later how these two observations, higher Δplasma:CSF and reduced CSF WBC counts, might be linked.
The second difference was in the rates of viral decay during the initial phase of therapy in CSF compared to plasma. Viral decay in CSF was slower than in plasma in subjects with lower CD4 counts, while decay rates in the two fluids were more often equal at higher CD4 counts. This finding extends our earlier report of this association [25]. It agrees, in part, with the findings of Ellis and colleagues [34] of a similar CD4 effect in a smaller series. However, this association of early decay differences with CD4 counts is at variance with a report by Eggers and colleagues [35] who correlated slower CSF decay after therapy with the presence of ADC or HIV encephalitis. Indeed, our results suggest that slower CSF decay was not confined to ADC subjects, but also occurred in neurologically asymptomatic subjects with low CD4 counts as noted by Ellis et al [34]. Further elucidation of these associations and the reasons for the minor differences in the experiences of different research groups will require a larger, more varied population sample, or combined analyses of the experience from several centers. Whatever the precise association, these observations suggest that the character of CSF infection changes with disease progression. In those with less advanced systemic disease, CSF infection often responds to potent ART as rapidly as blood infection. In those with more advanced systemic disease, with or without ADC, CSF HIV responds more slowly to antiviral therapy.
Overall effects of ART on CSF HIV infection
Before undertaking this study, we were concerned that CSF infection might not respond as well as systemic infection to treatment because of restricted penetration of many antiretroviral drugs [10]. This was not borne out by our observation that combination ART had a favorable impact on CSF HIV infection in the population of treated subjects. In the cross-sectional sample, both the plasma and CSF VLs were higher in untreated (median 4.93 and 3.61 log10 cpm in plasma and CSF, respectively) than in treated (median 4.00 and 1.66 log10 cpm in plasma and CSF) subjects (p < 0.001 for both variables compared by t-test). As noted also, HIV RNA was below the detection limit in a greater proportion of the CSF (38.9%) than plasma samples (18.5%). Given that the CSF HIV responded to therapy at least as effectively as plasma HIV, a relatively impaired CSF penetration of antiretroviral drugs seems unlikely to have influenced our findings. However, we did compare the CSF penetration of the antiretroviral regimens between those 'failing' to achieve virological suppression in plasma, and those with viral suppression, using a summed score in which each drug known to penetrate CSF counted as one and those not penetrating well counted as zero (data not shown). The means of the summed scores in the two groups (1.86 +/- 0.93, and 2.33 +/- 1.19, respectively) were not significantly different (p = 0.115). Of note, high CSF HIV was noted in rare treated subjects, sometimes exceeding plasma levels, as in subject 4034 who was treated with an unusual, poorly CSF-penetrating drug regimen [36] but responded well to an altered regimen. A second example of higher CSF than plasma HIV was subject 4032 who was suspected to be poorly adherent to his medications; genotypic testing supported this hypothesis by showing that his treatment failure was not due to drug resistance. Thus, although classified as on therapy, he likely was on intermittent and ineffective ART at best [37]. Overall, our experience suggests that effective CSF viral suppression is the general rule when plasma virus is suppressed.
The effectiveness of ART was clearly evident in the longitudinal treatment series. While the initial rate of CSF HIV decay lagged behind that of plasma in some subjects, all those with long-term follow-up achieved gratifying HIV reduction in this compartment with a time course generally proportional to their baseline level (Figure 3). None of the 28 patients showed persistent CSF virus in the face of undetectable plasma VL. The responses in subjects without previous treatment were similar to those reported by Polis and colleagues [30]with a protracted course but excellent suppression at 6 months. Our results also correlate well with the longer-term outcomes in more heterogeneous cohorts reported by others [38,39,35].
Effect of treatment of CSF infection in the presence of drug resistance
In addition to the generally favorable effects of ART on CSF infection, we observed an intriguing, salutary facet of treatment response. Namely, CSF infection showed a proportionally greater response to treatment than plasma infection in the face of antiviral drug resistance. Since drug resistance is the principal reason for long-term treatment failure, this may have very important implications for the prevention and treatment of CNS HIV infection and ADC. This superior effect of ART on CSF infection was noted in both the cross-sectional study and in the longitudinal treatment study. Moreover, it had been anticipated in our previous observations of CSF rebound in subjects with virological failure who underwent STI [18].
To examine this issue in the cross-sectional group, we compared the 36 subjects who were on therapy but had plasma VLs above 500 cpm (defined as treatment failures) with the 46 untreated subjects. While these two groups had similar plasma VLs (median 4.73 log10 cpm, IQR 3.99 – 5.04 in the treatment failure group, and 4.93 log10 cpm, IQR 4.53 – 5.48 in the untreated; p = 0.072 by t-test), their CSF HIV RNA levels differed substantially (2.45 log10 cpm median, IQR 1.45 – 3.64 log10 cpm in the treatment failure group, and 3.61 median, IQR 2.57 – 4.40 in the untreated; p = 0.002). As a result, the Δplasma:CSF was about ten-fold higher in the treatment failure (median, 2.19 log10 cpm, IQR 1.90 – 2.58) than in the untreated group (median 1.18 log10 cpm, IQR 0.42 – 2.16; p = 0.025); this 100-fold difference between plasma and CSF was similar to that reported by Stingele and colleagues in a study of paired specimens evaluating resistance mutations [29]. Thus, 'treatment failure' in the cross-sectional group was often associated with proportionally greater VL reductions in CSF than plasma.
This effect was further illustrated in the 'dissociated' CSF responses of the five treatment subjects who failed to clear virus in the plasma but responded completely or nearly completely to treatment in the CSF (Figure 5, Table 2 [see Additional file 1]). Their CSF HIV responses were both proportionally greater and more prolonged than those of plasma. Analysis of drug susceptibility in these 5 subjects showed that the main reason for continued plasma viremia was drug resistance. At least two patterns of responses were noted: 1) resistance at baseline leading to an initial small reduction in viremia and a new plateau (subjects 4001, 5001 and 5007, who switched medication at the start); and 2) a greater initial response followed by later partial recrudescence (subjects 4015 and 4030 who had been off therapy for some time after treatment failure before starting back on a new regimen). In the first two subjects, resistance was well established at baseline, while in the second two it emerged during the study, presumably from archived resistant strains [40]. Subject 5007 (Figure 5C) may have had a greater degree of compartmentalization of resistant virus. Genotypic testing revealed a plasma viral population with resistance-associated mutations at several amino acid sites (e.g., Protease mutation L90M and Reverse Transcriptase mutations M41L, K103N, Y181C, and T215Y). Testing of the CSF viral population revealed mixtures of wild-type with resistance-associated amino acids at all these sites. Such compartmentalized resistance with differences in drug susceptibility between CSF and plasma clearly occurs [41,29,42]. Notably, however, in the context of such systemic and CSF resistance, ART can be proportionally more effective in suppressing CSF infection than in treating systemic infection.
In fact, this greater effect on CSF than plasma in the face of drug resistance and treatment failure was predicted by observations on subjects failing ART who undertook STI [14,18]. In some of these subjects, the baseline CSF VLs were 10- to 1000-fold lower than baseline plasma VLs, but upon STI, rose to levels near or equal to those of plasma, indicating that the 'failed' therapy had suppressed the HIV to a greater extent in CSF than in plasma.
Models of blood-CSF virus and cell exchange and compartmentalization
To conceptualize CSF changes during infection, we framed 2 simple models of the exchange of HIV and lymphocytes between the blood and CNS compartments [9,14]. Figure 7 provides a schematic diagram of the elements of this model and examples of the variable relationships of CSF to plasma HIV noted in different clinical settings. The model invokes two basic types of infection: transitory and autonomous, as pictured in Panels A and B of Figure 7.
Figure 7 Model of CSF HIV infection. The diagram provides a simple schematic of hematogenous infection with T cells (including HIV-infected CD4 cells) and intrathecal macrophages separated by the blood-brain endothelial barrier. The model presumes that virus reaches the CNS principally within infected cells. T cells are shown as round cells, either infected (bar within nucleus) or uninfected (no bar). Similarly, the macrophages are shown as flat, elongated cells with or without infection (again, bar in nucleus). Both virus (circles with central dot) and cytokine/chemokine (smaller solid circles) are produced or provoked by infection on both sides of the barrier. Cells particularly involved in the illustrated process are highlighted in red and also may show thickened outline when active and broken line when the action is attenuated. Panels A, B, and C presents a simplified schematic of two basic types of CSF infection, transitory and autonomous, along with a combination of these types in mixed or amplified infection. Panels D-G apply these models to the relationships of plasma and CSF HIV (Δplasma:CSF) in four of the settings described in this report, including D. the high Δplasma:CSF in subjects with pleocytosis >10 cells/μL related to exuberant transitory infection; E the high Δplasma:CSF in ADC patients due to enhanced autonomous infection; F. the low Δplasma:CSF in subjects with < 50 blood CD4 cells/μL related to reduced transitory infection; and G the low Δplasma:CSF in treatment failures also related to decreased transitory infection.
Transitory infection (Figure 7A) refers to infection sustained by short-lived CD4+ T cells trafficking into the CSF space from the blood. In simplified terms, activation of lymphocytes outside of the nervous system favors their promiscuous entry [43,27], and when infected by HIV, these activated CD4 cells can release HIV into the surrounding fluid. This type of infection is determined principally from outside of the CNS, and depends upon systemic infection and cell activation. Infected and uninfected cells are pushed into the CSF.
Autonomous infection (Figure 7B) refers to infection that is sustained within the CNS by longer-lived cells and does not require continuous repletion from the blood. This is the type of infection that is also assumed to be 'compartmentalized' [44] and is likely sustained by longer-lived cells of the monocyte lineage that assume residence as perivascular and meningeal macrophages [45] rather than trafficking lymphocytes [46]. Whether such infection is fully self-sustaining or requires renewal from outside remains uncertain [47], but in it, the rates of turnover are lower and the cellular pools are different from those of transitory infection. When accompanied by CSF pleocytosis, the reactive cells are pulled into the CSF space [48]. Both of these basic types of infection may occur simultaneously or sequentially, and indeed mixed infection with varying contributions of the two types may be the rule (Figure 7C).
We hypothesize that transitory infection predominates during the early phases of systemic HIV infection but gives way to an increasing component of autonomous infection as immune deregulation progresses. One reason for hypothesizing the early predominance of transitory infection is that the rapid, early-phase decay of CSF HIV RNA noted in some subject initiating therapy is equivalent to that occurring in the plasma. The rate of plasma HIV decay depends upon the potency of the antiviral regimen [49], and as drug exposure in the CNS is almost always lower than at most systemic sites (by virtue of the blood-brain and blood-CSF barriers and other factors reducing drug penetration), we would expect the potency of the regimen in CSF VL to be lower. However, incorporating a model of transitory CSF infection, suppression of systemic infection would also lower the infection rate of trafficking lymphocytes and virus-induced activation, thereby reducing the influx of these cells and rapidly decreasing the CSF VL. Evolution from transitory to more autonomous infection at later stages of disease would explain the slower CSF decay relative to plasma, as decay kinetics are now closely associated with reduction in T cells. CSF infection may also be amplified by entry of activated infected or target CD4 cells which increase their transit across the endothelial barrier in response to chemokines and other signals (Panel C of Figure 7). Amplified infection might obscure the underlying importance of autonomous infection as the inciting process, and might respond more readily to systemic treatment.
Variations of these models can also be invoked to explain the relationships of CSF to plasma VL noted in some of the subject groups in this study. For example, the two situations where we found smaller Δplasma:CSF. values might relate to quite different mechanisms. Specifically, in the first instance the association of pleocytosis (>10 cells/μL) with high CSF HIV RNA (approaching that of the plasma VL with both ≥ 10,000 cpm), and small Δplasma:CSF might relate to exuberant cell entry and a high level of transitory or amplified infection as diagrammed in Figure 7D. This mechanism would also account for high CSF HIV concentrations in patients with meningitis due to other infections such as Mycobacterium tuberculosis and Cryptococcus neoformans [4,50], where cell entry is driven by an unrelated process but result in local amplification of CSF infection. In the second instance, narrow or even reversed (negative) Δplasma:CSF is best explained by a high level of autonomous infection within perivascular or parenchymal macrophages in patients with ADC (Figure 7E).
Figure 7 also illustrates hypothetical explanations for the increased Δplasma:CSF noted in two settings: subjects with CD4 counts < 50 cells/μL (F) and those with treatment failure (G). In both cases, a diminished transitory component of infection may be involved, either because of insufficient CD4 target cells or reduced signals and responses to cell activation. In the case of low CD4 cells, the reduced transitory component uncovers low-level autonomous CSF infection. Of course, comparison of this setting with ADC in which CD4 cells are often decreased underscores a switch from low-level and seemingly benign autonomous infection to more active and 'malignant' macrophage infection with resultant toxic sequelae. This then raises the question of what causes the change in the character of autonomous infection in this setting – a change in the virus population or an alteration in the host? The increased Δplasma:CSF in treatment failure is supported by observations of diminished cell activation in this setting [51], but is this sufficient to reduce CSF HIV to the extent observed? Might reduced activation and lower CSF HIV levels relate to reduced fitness of the resistant viruses [52] or simply to a quantitative decrement in systemic infection?
These considerations also do not fully preclude the importance of intrathecal drug effects. Simply considering extracellular drug levels in CSF compared to plasma may not be sufficient. Intracellular drug activity of nucleoside RT inhibitors might be enhanced within the cells supporting autonomous CNS infection despite lower extracellular drug concentrations than in systemic sites of infection [53]. Subject 4034, with disproportionate virological failure in CSF compared to blood in the face of a poorly penetrating ART regimen, illustrates the potential importance of drug penetration in the presence of autonomous infection. His high CSF VL was remedied by a regimen with greater penetration.
Neurological implications
These studies have relevance for two clinical neurological disorders related to CNS HIV infection: aseptic meningitis and ADC. The first is a component of the central focus of the study, while the second stands as an important background issue.
Aseptic or HIV meningitis is the clinical diagnosis conferred on HIV-infected patients with CSF pleocytosis without alternative cause, irrespective of clinical symptoms or signs such as headache, photophobia and stiff neck [54-56]. We extend the characterization of this aseptic meningitis, defining its virological profile, clinical context and response to therapy. In the cross-sectional sample, CSF pleocytosis was detected in those with CD4 counts above 50 cells/μL, and was associated with plasma VLs near or above 10,000 cpm.
The frequency of pleocytosis in our subjects and their lack of symptoms confirms earlier observations reported at a time when HIV infection was less well characterized [2,3]. Our subjects were queried and assiduously examined for symptoms or signs of meningitis, and tested for associated abnormality or changes in QNPZ-4. None were found, and we were unable to predict the presence of elevated WBC counts from interviews or examinations. Why is HIV-associated pleocytosis asymptomatic when other infections with similar CSF cell counts are often accompanied by clinical symptoms and signs? Presumably, there are differences in the character of the cells, and more particularly in their secreted products, that determine the absence or presence of symptoms.
Since CSF pleocytosis is frequent and characteristically asymptomatic, how does one approach the HIV-infected patient presenting with headache who has an elevated CSF lymphocyte count [56]? When is pleocytosis incidental and headache due to another cause? When can pleocytosis be attributed to the underlying HIV infection and thus not warrant further diagnostic evaluations? Our experience allows some initial answers to these questions. First, pleocytosis related only to HIV is uncommon in those with blood CD4 counts below 50 cells/μL, and therefore an increased CSF cell count in this setting should be suspect. Second, since ART usually eliminates elevated WBC counts, pleocytosis in a treated patient should be regarded similarly, with the additional consideration of adherence to therapy. Third, since headache is rarely provoked by simple HIV-related pleocytosis, a search for another type of CNS infection or process is warranted in such a case. Finally, in chronic pleocytosis without headache, a trial of ART may confirm HIV as the cause if the cell count resolves. Perhaps the term aseptic meningitis in this setting should be reserved for those with symptoms or signs accompanying CSF abnormalities. The larger group of individuals with clinically silent elevated CSF cell counts may simply be designated as having asymptomatic pleocytosis of HIV infection.
ADC is considered to be caused by brain HIV infection, but mediated by 'indirect' mechanisms involving a variety of pathways initiated and sustained by viral and host gene products [57-61]. Infected and uninfected macrophages and perhaps microglia appear to play a central role in these processes [45]. The current study focused on CSF HIV infection and was not designed to more directly assess the relationship of parenchymal brain HIV infection to ADC. Nonetheless, it afforded a view of certain aspects of this important issue.
Neurological performance was assessed using the QNPZ-4 score. This correlated highly with the presence and stage of ADC; the reduced QNPZ-4 score (median -0.50) of the cross-sectional sample was largely the result of including the ADC subjects. Longitudinally, the QNPZ-4 score afforded a stable measure in those without change in clinical neurological status, while it tracked a pattern of improvement in ADC subjects responding to ART that mirrored their antiviral response in CSF and plasma (Figure 6). Cross-sectionally, both ADC stages and QNPZ-4 scores correlated with blood CD4 counts, though neither correlated overall with plasma or CSF HIV RNA levels. The association of ADC with HIV progression and depressed CD4 blood counts has been well documented [62], and the lack of correlation of neurological disease with the magnitude of systemic or CSF infection over the broad range of CD4 counts has also been previously noted [6,7]. Our findings agree with some [8] but differ from other reports [7] in that we also did not find a strong correlation between the neurological and virological measurements in subjects with lower CD4 counts. There was only a weak correlation between CSF VL and ADC stage (p = 0.037, rho = 0.316) in subjects with CD4 counts below 200 cells/μL, and no significant correlation with QNPZ-4 score in this group; no correlation of the CSF HIV concentration with ADC stage or QNPZ-4 score was found in subjects with CD4 counts <50 cells/μL. Moreover, we found no correlation between the neurological measures and CSF VLs among the entire group of untreated subjects, or in those with CD4 counts below 200 cells/μL.
This does not mean that HIV brain infection is not central to the pathogenesis of ADC or that HIV brain infection cannot be reflected in CSF, but only that such an association is obscured by other factors. In this study, there may have been two principal reasons why no association was found. First, there was a high 'background' of increased CSF HIV in neurologically normal subjects. Second, many of our ADC subjects were on ART and clinically stable, and therefore likely suffered 'inactive' or residual brain disease. The lack of diagnostic specificity of CSF HIV RNA measurements underscores the varied relationship between CSF changes, on the one hand, and brain infection and disease, on the other. The CSF space and brain parenchyma are best viewed as two separate but intersecting compartments in which infection is not always congruent, necessitating caution in interpreting brain events from CSF findings. At least three types of intersections may occur. First, brain infection can 'overflow' into the CSF so that brain-derived HIV (and other markers) can be detected and directly measured in this fluid. The likely major cells of origin for HIV in these cases of autonomous infection are perivascular macrophages and perhaps the parenchymal macrophages and microglia [45], with virus reaching the CSF by diffusion along the perivascular spaces. CSF in this setting can provide a direct sample of brain infection. Second, CSF and brain infection may be parallel, although not necessarily identical. Because the leptomeninges and brain are both non-lymphatic organs and separated from the blood by barriers to the free passage of viruses, immune defenses (both humoral and cellular) and drugs, infection and host responses may be sufficiently similar in the meninges and brain to permit CSF analysis to monitor brain infection, whether autonomous or transitory. Third, infections might markedly diverge in these compartments, in which case CSF would provide little direct insight into brain infection; at the extreme, the CSF might simply reflect meningeal infection which has no counterpart within the brain. The problem for the clinician is to distinguish which of these three relationships predominates in a given patient.
Measuring the VL in CSF is not enough to make these important distinctions, and the CSF HIV RNA concentration cannot, in isolation, serve as a reliable diagnostic marker for ADC. Nor can the presence of pleocytosis be used to rule in or out ADC. This is unfortunate since objective markers are needed both in the clinic for practical diagnosis and in the research setting to more precisely define therapeutic targets. In the occasional patient, such as those designated as outliers in Figure 2, substantially higher HIV RNA in CSF than in plasma may be diagnostically suggestive, but in most cases neither the level of CSF virus nor its relation to the plasma VL distinguishes those with ADC. Future efforts need to assess the value of supplementary CSF measurements, involving more detailed characterization of the virus (cell tropism, chemokine co-receptor utilization, or still-elusive markers of neuropathogenicity) [63-65] along with the use of ancillary markers of immune responses and neural injury [8,66,67].
An additional clinical need is for laboratory measures that assess the activity of ADC and underlying HIV encephalitis. While we included subjects in this study with treated ADC and suppressed CSF infection, this would not be appropriate in a clinical trial assessing ADC treatment. Measuring CSF HIV concentration may be helpful in this setting, since undetectable CSF HIV likely signals suppressed brain infection.
Conclusion
HIV infection of the CNS is a nearly ubiquitous facet of systemic infection, but varies in character and clinical consequences. From very early exposure during primary systemic infection, most HIV-infected patients experience chronic asymptomatic CNS infection. However, a few individuals will develop encephalitis presenting as ADC. CSF sampling provides a valuable window into this infection and its variability.
We have framed the discussion of our results in models of CSF infection. Embedded in these models are a number of dichotomies relating systemic to CNS infections: (1) transitory versus more autonomous infections with rapid versus slower turnover rates; (2) CSF lymphocytosis either causing or responding to local infection, and (3) infection of the meninges presenting as 'asymptomatic pleocytosis' versus more toxic parenchymal or perivascular infection leading to the brain dysfunction of ADC. These provide a framework for future studies examining the mechanisms of infection in molecular terms and with respect to cell and HIV exchange and compartmentalization.
On a more practical level, to the extent that CSF infection reflects infection of the brain, antiretroviral therapy is usually effective in suppressing CNS HIV replication. Our longitudinal observations show that CSF infection usually responds well to combination antiretroviral therapy, equaling or exceeding systemic responses as reflected in plasma. Even where resistance leads to virological failure and persistent plasma viremia, ART may have a salutary effect on CSF. While the mechanisms underlying these favorable treatment effects remain uncertain, these observations are consonant with other reports using less frequent monitoring and are therapeutically reassuring. Our findings suggest that favorable virological outcomes in the CSF are the rule rather than an exception. They are also consistent with clinical studies that report a falling incidence of ADC in the current treatment era [68].
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
SS participated in analysis and interpretation of the data and helped to draft the manuscript. AN helped with subject evaluations. NL participated in analysis and preparation of the data. TL helped in the study planning and coordinated the viral load assays. CP participated in planning of the genotypic and phenotypic resistance studies. SD participated in the study planning and design. EP participated in planning and interpretation of the genotypic and phenotypic resistance studies, and helped prepare the manuscript. RWP conceived of the design and analysis of the study and drafted the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Supplementary Material
Additional File 1
Table 2. Clinical data, ART regimen history, and genotypic analysis of resistance mutations from five subjects with dissociated treatment responses. Antiretroviral medications able to penetrate into the CSF [10] are indicated in bold font. Mutations are given as differences from a drug-sensitive control (e.g., NL 4–3). Wild-type amino acids are indicated by the single capitalized letters in black while substitutions are indicated by either red or blue font. In red, but not bold, are minor mutations associated with resistance to a drug the subject was taking at the time of sampling. However, red and bolded type indicates major mutations associated with resistance to a drug that the subject was taking at the time of sampling. In blue, not bold, are minor mutations associated with resistance to a drug that the subject was not currently taking, while blue-bold typeface indicates major mutations associated with resistance to a drug which subject was not taking at the time.
Click here for file
Acknowledgements
This work was supported by: NIH grants R01 NS37660, R01 MH62701, and UCSF GCRC, 5-MO1-RR-00083, Abacavir was provided by Glaxo-Wellcome. We would also like to thank the subjects who volunteered for these studies and the staff of the SFGH/UCSF GCRC clinical facility and Virology Laboratory for their invaluable help.
==== Refs
Navia BA Jordan BD Price RW The AIDS dementia complex: I. Clinical Features Ann Neurol 1986 19 517 524 3729308 10.1002/ana.410190602
Marshall DW Brey RL Cahill WT Houk RW Zajac RA Boswell RN Spectrum of cerebrospinal fluid findings in various stages of human immunodeficiency virus infection Arch Neurol 1988 45 954 958 2970837
McArthur JC Cohen BA Farzadegan H Cornblath DR Selnes OA Ostrow D Johnson RT Phair J Polk BF Cerebrospinal fluid abnormalities in homosexual men with and without neuropsychiatric findings Ann Neurol 1988 23(suppl) S34 S37 3348599 10.1002/ana.410230712
Brew BJ Pemberton L Cunningham P Law MG Levels of human immunodeficiency virus type 1 RNA in cerebrospinal fluid correlate with AIDS dementia stage J Inf Dis 1997 175 963 966 9086160
Pilcher CD Shugars DC Fiscus SA Miller WC Menezes P Giner J Dean B Robertson K Hart CE Lennox JL Eron JJJ Hicks CB HIV in body fluids during primary HIV infection: implications for pathogenesis, treatment and public health Aids 2001 15 837 845 11399956 10.1097/00002030-200105040-00004
Ellis RJ Hsia K Spector SA Nelson JA Heaton RK Wallace MR Abramson I Atkinson JH Grant I McCutchan JA Cerebrospinal fluid human immunodeficiency virus type 1 RNA levels are elevated in neurocognitively impaired individuals with acquired immunodeficiency syndrome. HIV Neurobehavioral Research Center Group [see comments] Ann Neurol 1997 42 679 688 9392566 10.1002/ana.410420503
McArthur JC McClernon DR Cronin MF Nance-Sproson TE Saah AJ St Clair M Lanier ER Relationship between human immunodeficiency virus-associated dementia and viral load in cerebrospinal fluid and brain [see comments] Ann Neurol 1997 42 689 698 9392567 10.1002/ana.410420504
Gisslen M Fuchs D Svennerholm B Hagberg L Cerebrospinal fluid viral load, intrathecal immunoactivation, and cerebrospinal fluid monocytic cell count in HIV-1 infection J Acquir Immune Defic Syndr 1999 21 271 276 10428104
Price RW The two faces of cerebrospinal fluid (CSF) HIV infection Trends in Microbiology 2000
Enting RH Hoetelmans RMW Lange JMA Burger DM Beijnen JH Portegies P Antiretroviral drugs and the central nervous system AIDS 1998 12 1941 1955 9814862
Lambotte O Deiva K Tardieu M HIV-1 persistence, viral reservoir, and the central nervous system in the HAART era Brain Pathol 2003 13 95 103 12580549
Antinori A Cingolani A Giancola ML Forbici F De Luca A Perno CF Clinical implications of HIV-1 drug resistance in the neurological compartment Scand J Infect Dis Suppl 2003 35 Suppl 106 41 44 15000582 10.1080/03008870310009650
Price RW Staprans S Measuring the "viral load" in cerebrospinal fluid in human immunodeficiency virus infection: window into brain infection? [editorial; comment] Ann Neurol 1997 42 675 678 9392565 10.1002/ana.410420502
Price RW Deeks SG Antiretroviral drug treatment interruption in human immunodeficiency virus-infected adults: Clinical and pathogenetic implications for the central nervous system J Neurovirol 2004 10 Suppl 1 44 51 14982739 10.1080/13550280490268223
Price RW Hoh R Drews B Deeks S Grant R Changes in cerebrospinal fluid (CSF) after interruption of antiretroviral therapy (ART): Accelerated rise in CSF HIV-1 RNA and accompanying CSF pleocytosis.: ; San Francisco. 2000 306 10709039
Daluge SM Good SS Faletto MB Miller WH St Clair MH Boone LR Tisdale M Parry NR Reardon JE Dornsife RE Averett DR Krenitsky TA 1592U89, a novel carbocyclic nucleoside analog with potent, selective anti-human immunodeficiency virus activity Antimicrob Agents Chemother 1997 41 1082 1093 9145874
Lanier ER Sturge G McClernon D Brown S Halman M Sacktor N McArthur J Atkinson JH Clifford D Price RW Simpson D Torres G Catalan J Marder K Power C Hall C Romero C Brew B HIV-1 reverse transcriptase sequence in plasma and cerebrospinal fluid of patients with AIDS dementia complex treated with Abacavir Aids 2001 15 747 751 11371689 10.1097/00002030-200104130-00010
Price RW Paxinos EE Grant RM Drews B Nilsson A Hoh R Hellmann NS Petropoulos CJ Deeks SG Cerebrospinal fluid response to structured treatment interruption after virological failure Aids 2001 15 1251 1259 11426069 10.1097/00002030-200107060-00006
Price RW Sidtis JJ Evaluation of the AIDS dementia complex in clinical trials J AIDS 1990 3(Supp (2)) S51 S60
Sidtis JJ Evaluation of the AIDS dementia complex in adults Res Publ Assoc Res Nerv Ment Dis 1994 72 273 287 8115718
Price RW Management of AIDS dementia complex and HIV-1 infection of the nervous system Aids 1995 9 Suppl A S221 36 8819589
Price RW Yiannoutsos CT Clifford DB Zaborski L Tselis A Sidtis JJ Cohen B Hall CD Erice A Henry K Neurological outcomes in late HIV infection: adverse impact of neurological impairment on survival and protective effect of antiviral therapy. AIDS Clinical Trial Group and Neurological AIDS Research Consortium study team Aids 1999 13 1677 1685 10509569 10.1097/00002030-199909100-00011
Nomenclature and research case definitions for neurologic manifestations of human immunodeficiency virus-type 1 (HIV-1) infection. Report of a Working Group of the American Academy of Neurology AIDS Task Force Neurology 1991 41 778 785 2046917
Sidtis JJ Price RW Early HIV-1 infection and the AIDS dementia complex Neurology 1990 40 323 326 2300258
Staprans S Inkina N Glidden D Novakovic-Agopian T Grant RM Heyes M Aweeka F Deeks S Price RW Time Course of Cerebrospinal Fluid (CSF) Responses to Antiretroviral Therapy: Evidence for Variable Compartmentalization of Infection AIDS 1999 13 1051 1061 10397535 10.1097/00002030-199906180-00008
Petroupoulos CJ Parkin NT Limoli KL Lie YS Wrin T Huang W Tian H Smith D Winslow GA Capon DJ Whitcomb JM A novel phenotypic drug suseptibility assay for human immunodeficiency virus type 1 Antimicrob Agents Chemother 2000 44 920 928 10722492 10.1128/AAC.44.4.920-928.2000
Ransohoff RM Kivisakk P Kidd G Three or more routes for leukocyte migration into the central nervous system Nat Rev Immunol 2003 3 569 581 12876559 10.1038/nri1130
Gisslen M Hagberg L Fuchs D Norkrans G Svennerholm B Cerebrospinal fluid viral load in HIV-1-infected patients without antiretroviral treatment: a longitudinal study J Acquir Immune Defic Syndr Hum Retrovirol 1998 17 291 295 9525428
Stingele K Haas J Zimmermann T Stingele R Hubsch-Muller C Freitag M Storch-Hagenlocher B Hartmann M Wildemann B Independent HIV replication in paired CSF and blood viral isolates during antiretroviral therapy Neurology 2001 56 355 361 11171901
Polis MA Suzman DL Yoder CP Shen JM Mican JM Dewar RL Metcalf JA Falloon J Davey RTJ Kovacs JA Feinberg MB Masur H Piscitelli SC Suppression of cerebrospinal fluid HIV burden in antiretroviral naive patients on a potent four-drug antiretroviral regimen Aids 2003 17 1167 1172 12819518 10.1097/00002030-200305230-00008
Ellis RJ Childers ME Zimmerman JD Frost SD Deutsch R McCutchan JA Human immunodeficiency virus-1 RNA levels in cerebrospinal fluid exhibit a set point in clinically stable patients not receiving antiretroviral therapy J Infect Dis 2003 187 1818 1821 12751041 10.1086/375152
Martin C Albert J Hansson P Pehrsson P Link H Sonnerborg A Cerebrospinal fluid mononuclear cell counts influence CSF HIV-1 RNA levels J Acquir Immun Def Syndr Human Retrovirol 1998 17 214 219
Garcia F Niebla G Romeu J Vidal C Plana M Ortega M Ruiz L Gallart T Clotet B Miro JM Pumarola T Gatell JM Cerebrospinal fluid HIV-1 RNA levels in asymptomatic patients with early stage chronic HIV-1 infection: support for the hypothesis of local virus replication AIDS 1999 13 1491 1496 10465072 10.1097/00002030-199908200-00008
Ellis RJ Gamst AC Capparelli E Spector SA Hsia K Wolfson T Abramson I Grant I McCutchan JA Cerebrospinal fluid HIV RNA originates from both local CNS and systemic sources Neurology 2000 54 927 936 10690988
Eggers C Hertogs K Sturenburg HJ van Lunzen J Stellbrink HJ Delayed central nervous system virus suppression during highly active antiretroviral therapy is associated with HIV encephalopathy, but not with viral drug resistance or poor central nervous system drug penetration Aids 2003 17 1897 1906 12960822 10.1097/00002030-200309050-00008
De Luca A Ciancio BC Larussa D Murri R Cingolani A Rizzo MG Giancola ML Ammassari A Ortona L Correlates of independent HIV-1 replication in the CNS and of its control by antiretrovirals Neurology 2002 59 342 347 12177366
Bangsberg DR Charlebois ED Grant RM Holodniy M Deeks SG Perry S Conroy KN Clark R Guzman D Zolopa A Moss A High levels of adherence do not prevent accumulation of HIV drug resistance mutations Aids 2003 17 1925 1932 12960825 10.1097/00002030-200309050-00011
Antinori A Giancola ML Grisetti S Soldani F Alba L Liuzzi G Amendola A Capobianchi M Tozzi V Perno CF Factors influencing virological response to antiretroviral drugs in cerebrospinal fluid of advanced HIV-1-infected patients Aids 2002 16 1867 1876 12351946 10.1097/00002030-200209270-00003
Bestetti A Presi S Pierotti C Bossolasco S Sala S Racca S Carrera P Lazzarin A Cinque P Long-term virological effect of highly active antiretroviral therapy on cerebrospinal fluid and relationship with genotypic resistance J Neurovirol 2004 10 Suppl 1 52 57 14982740
Siliciano JD Siliciano RF A long-term latent reservoir for HIV-1: discovery and clinical implications J Antimicrob Chemother 2004 54 6 9 15163657 10.1093/jac/dkh292
Cunningham P Smith D Satchell C Cooper DA Brew BJ Evidence for independent development of reverse transcriptase inhibitor resistance patterns in cerebrospinal fluid AIDS 2000
Cinque P Presi S Bestetti A Pierotti C Racca S Boeri E Morelli P Carrera P Ferrari M Lazzarin A Effect of genotypic resistance on the virological response to highly active antiretroviral therapy in cerebrospinal fluid AIDS Res Hum Retroviruses 2001 17 377 383 11282006 10.1089/088922201750102409
Hickey WF Lassmann H Cross AH Keane RW and Hickey WF Lymphocyte entry and the initiation of inflammation in the central nervous system Immunology of the Nervous System 1997 New York, Oxford University Press 200 225
Strain MC Letendre S Pillai SK Russell T Ignacio CC Gunthard HF Good B Smith DM Wolinsky SM Furtado M Marquie-Beck J Durelle J Grant I Richman DD Marcotte T McCutchan JA Ellis RJ Wong JK Genetic composition of human immunodeficiency virus type 1 in cerebrospinal fluid and blood without treatment and during failing antiretroviral therapy J Virol 2005 79 1772 1788 15650202 10.1128/JVI.79.3.1772-1788.2005
Williams KC Hickey WF Central nervous system damage, monocytes and macrophages, and neurological disorders in AIDS Annu Rev Neurosci 2002 25 537 562 12052920 10.1146/annurev.neuro.25.112701.142822
Haas DW Johnson BW Spearman P Raffanti S Nicotera J Schmidt D Hulgan T Shepard R Fiscus SA Two phases of HIV RNA decay in CSF during initial days of multidrug therapy Neurology 2003 61 1391 1396 14638961
Gartner S HIV infection and dementia Science 2000 287 602 604 10691542 10.1126/science.287.5453.602
Shacklett BL Cox CA Wilkens DT Karl Karlsson R Nilsson A Nixon DF Price RW Increased adhesion molecule and chemokine receptor expression on CD8+ T cells trafficking to cerebrospinal fluid in HIV-1 infection J Infect Dis 2004 189 2202 2212 15181567 10.1086/421244
Markowitz M Louie M Hurley A Sun E Di Mascio M Perelson AS Ho DD A novel antiviral intervention results in more accurate assessment of human immunodeficiency virus type 1 replication dynamics and T-cell decay in vivo J Virol 2003 77 5037 5038 12663814 10.1128/JVI.77.8.5037-5038.2003
Morris L Silber E Sonnenberg P Eintracht S Nyoka S Lyons SF Saffer D Koornhof H Martin DJ High human immunodeficiency virus type 1 RNA load in the cerebrospinal fluid from patients with lymphocytic meningitis J Infectious Diseases 1998 177 473 476 9466541
Deeks SG Barbour JD Grant RM Martin JN Duration and predictors of CD4 T-cell gains in patients who continue combination therapy despite detectable plasma viremia Aids 2002 16 201 207 11807304 10.1097/00002030-200201250-00009
Barbour JD Hecht FM Wrin T Segal MR Ramstead CA Liegler TJ Busch MP Petropoulos CJ Hellmann NS Kahn JO Grant RM Higher CD4+ T cell counts associated with low viral pol replication capacity among treatment-naive adults in early HIV-1 infection J Infect Dis 2004 190 251 256 15216458 10.1086/422036
Foudraine NA Hoetelmans RMW Lange JMA deWolf F vanBenthem BHB Maas JJ Keet IPM Portegies P Cerebrospinal-fluid HIV-1 RNA and drug concentrations after treatment with lamivudine plus zidovudine or stavudine Lancet 1998 351 1547 1551 10326538 10.1016/S0140-6736(98)07333-4
Hollander H Stringari S Human immunodeficiency virus-associated meningitis: Clinical course and correlations Am J Med 1987 83 813 816 3674088 10.1016/0002-9343(87)90635-8
Brew BJ Miller J Human immunodeficiency virus-related headache Neurology 1993 43 1098 1100 8170549
Hollander H McGuire D Burack JH Diagnostic lumbar puncture in HIV-infected patients: analysis of 138 cases Am J Med 1994 96 223 228 7908777 10.1016/0002-9343(94)90146-5
Krebs FC Ross H McAllister J Wigdahl B HIV-1-associated central nervous system dysfunction Adv Pharmacol 2000 49 315 385 11013768
Kaul M Garden GA Lipton SA Pathways to neuronal injury and apoptosis in HIV-associated dementia Nature 2001 410 988 994 11309629 10.1038/35073667
Turchan J Sacktor N Wojna V Conant K Nath A Neuroprotective therapy for HIV dementia Curr HIV Res 2003 1 373 383 15049424 10.2174/1570162033485113
Nath A Berger J HIV Dementia Curr Treat Options Neurol 2004 6 139 151 14759346
Gonzalez-Scarano F Martin-Garcia J The neuropathogenesis of AIDS Nat Rev Immunol 2005 5 69 81 15630430 10.1038/nri1527
McArthur JC Hoover DR Bacellar H Miller EN Cohen BA Becker JT Graham NM McArthur JH Selnes OA Jacobson LP Dementia in AIDS patients: incidence and risk factors. Multicenter AIDS Cohort Study Neurology 1993 43 2245 2252 8232937
Power C McArthur JC Johnson RT Griffin DE Glass JD Dewey R Chesebro B Distinct HIV-1 env sequences are associated with neurotropism and neurovirulence Curr Top Microbiol Immunol 1995 202 89 104 7587373
Hogan TH Stauff DL Krebs FC Gartner S Quiterio SJ Wigdahl B Structural and functional evolution of human immunodeficiency virus type 1 long terminal repeat CCAAT/enhancer binding protein sites and their use as molecular markers for central nervous system disease progression J Neurovirol 2003 9 55 68 12587069
Power C Zhang K van Marle G Comparative neurovirulence in lentiviral infections: The roles of viral molecular diversity and select proteases J Neurovirol 2004 10 Suppl 1 113 117 14982749 10.1080/13550280490270815
Hagberg L Fuchs D Rosengren L Gisslen M Intrathecal immune activation is associated with cerebrospinal fluid markers of neuronal destruction in AIDS patients J Neuroimmunol 2000 102 51 55 10626666 10.1016/S0165-5728(99)00150-2
Abdulle S Hagberg L Svennerholm B Fuchs D Gisslen M Continuing intrathecal immunoactivation despite two years of effective antiretroviral therapy against HIV-1 infection Aids 2002 16 2145 2149 12409735 10.1097/00002030-200211080-00006
d'Arminio Monforte A Cinque P Mocroft A Goebel FD Antunes F Katlama C Justesen US Vella S Kirk O Lundgren J Changing incidence of central nervous system diseases in the EuroSIDA cohort Ann Neurol 2004 55 320 328 14991809 10.1002/ana.10827
|
16266436
|
PMC1299327
|
CC BY
|
2021-01-04 16:28:16
|
no
|
BMC Infect Dis. 2005 Nov 2; 5:98
|
utf-8
|
BMC Infect Dis
| 2,005 |
10.1186/1471-2334-5-98
|
oa_comm
|
==== Front
Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-1341627765910.1186/1465-9921-6-134ResearchThe Beijing genotype and drug resistant tuberculosis in the Aral Sea region of Central Asia Cox Helen Suzanne [email protected] Tanja [email protected] Daribay [email protected] Yared [email protected]üsch-Gerdess Sabine [email protected] Stefan [email protected] Médecins Sans Frontières (MSF), Aral Sea Area Programme, Uzbekistan and Turkmenistan Tashkent, Uzbekistan2 National Reference Center for Mycobacteria, Forschungszentrum Borstel, Borstel, Germany3 Ministry of Health, Nukus, Karakalpakstan, Uzbekistan4 Médecins Sans Frontières (MSF), Amsterdam, Holland2005 8 11 2005 6 1 134 134 25 8 2005 8 11 2005 Copyright © 2005 Cox et al; licensee BioMed Central Ltd.2005Cox et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
After the collapse of the Soviet Union, dramatically increasing rates of tuberculosis and multidrug-resistant tuberculosis (MDR-TB) have been reported from several countries. This development has been mainly attributed to the widespread breakdown of TB control systems and declining socio-economic status. However, recent studies have raised concern that the Beijing genotype of Mycobacterium tuberculosis might be contributing to the epidemic through its widespread presence and potentially enhanced ability to acquire resistance.
Methods
A total of 397 M. tuberculosis strains from a cross sectional survey performed in the Aral Sea region in Uzbekistan and Turkmenistan have been analysed by drug susceptibility testing, IS6110 fingerprinting, and spoligotyping.
Results
Fifteen isolates showed mixed banding patterns indicating simultaneous infection with 2 strains. Among the remaining 382 strains, 152 (40%) were grouped in 42 clusters with identical fingerprint and spoligotype patterns. Overall, 50% of all isolates were Beijing genotype, with 55% of these strains appearing in clusters compared to 25% of non-Beijing strains. The percentage of Beijing strains increased with increasing drug resistance among both new and previously treated patients; 38% of fully-susceptible isolates were Beijing genotype, while 75% of MDR-TB strains were of the Beijing type.
Conclusion
The Beijing genotype is a major cause of tuberculosis in this region, it is strongly associated with drug resistance, independent of previous tuberculosis treatment and may be strongly contributing to the transmission of MDR-TB. Further investigation around the consequences of Beijing genotype infection for both tuberculosis transmission and outcomes of standard short course chemotherapy are urgently needed.
==== Body
Background
Tuberculosis (TB) remains one of the leading infectious killers worldwide, with an estimated 2 million deaths annually (1). In the year 2002 the number of incident TB cases was estimated at 8.8 million (2). Globally there is a 1.8% annual rise in new tuberculosis cases, with a 6% yearly increase in the former Soviet Union (1).
These data are increasingly accompanied by the phenomenon of drug-resistance, making successful treatment and control of the disease even more difficult. The third report on global surveillance for tuberculosis drug-resistance reveals alarming levels of MDR-TB (resistance at least to isoniazid and rifampicin) of up to 14% among new cases, with an estimated 300,000 new cases of MDR-TB globally per year [3]. Of particular concern are parts of Eastern Europe and Central Asia where tuberculosis patients are 10 times more likely to have MDR-TB than in the rest of the world [4].
The treatment of patients with MDR-TB is extremely difficult, expensive, and requires special treatment regimens and case management [5]. In addition, patients infected with MDR-TB may remain infectious for prolonged periods of times further accelerating the spread of MDR-TB. It is therefore important to understand factors contributing to the development of MDR-TB and the potential epidemiological impact of MDR-TB strains in order to develop effective control strategies.
Increasing tuberculosis incidence and the emergence of MDR-TB in the former Soviet Union have been mainly attributed to the deterioration of economic and social conditions as well as to the widespread breakdown of tuberculosis control systems since the late 1980s [6,7]. The possibility that the pathogen itself is also contributing to this problem has been recently suggested by two studies from the Russian Federation, which found high proportions of a particular genotype of tuberculosis, namely the Beijing genotype, which was strongly associated with drug resistance [8,9].
This notion has been further supported by recent studies that have provided evidence that the genetic heterogeneity of Mycobacterium tuberculosis complex isolates is greater than previously thought, and might influence the transmissibility and virulence of particular isolates [10-13]. Strains of the Beijing genotype were first described in China and neighbouring countries in 1995 [14], and subsequently the occurrence of Beijing genotype strains has been documented in several parts of the world [8,9,14-17]. The Beijing genotype has caused outbreaks of MDR-TB [18,19], and some, but not all studies, indicate an association with drug resistance [16]. However, there is limited information available on both the occurrence and effects of the Beijing genotype, especially from areas with high tuberculosis incidence and high rates of MDR-TB. If these strains have an enhanced capacity to gain resistance, this will have serious consequences for the treatment of tuberculosis.
In a recent study, we found high levels of MDR-TB in the Aral Sea region in Uzbekistan and Turkmenistan, Central Asia [20]. A cross-sectional survey of more than 400 smear-positive tuberculosis patients, revealed levels of MDR-TB of 27% in Karakalpakstan (Uzbekistan) and 11% in Dashoguz (Turkmenistan). The DOTS strategy was introduced progressively into this region from 1998 and now covers a population of around 4 million people. High case notification rates for smear positive tuberculosis: 190/100,000/year in Karakalpakstan and 70/100,000/year in Dashoguz are reported from the DOTS programme.
In this study, we elucidate the importance of the Beijing genotype for tuberculosis in the Aral Sea region. Molecular typing of the isolates from the drug resistance survey was performed to determine the proportion of patients infected with Beijing genotype strains, and associations with drug resistance and other patient characteristics were analysed.
Materials and methods
Study population
A cross-sectional survey for anti-tuberculosis drug-resistance was conducted in 4 districts in the Autonomous Republic of Karakalpakstan, Uzbekistan, and in 4 districts in Dashoguz Velayat, Turkmenistan. Smear positive pulmonary tuberculosis patients initiating DOTS treatment in these districts were included in the study. The study was based on the recommendations for drug resistance surveys outlined by WHO and IUATLD [21]. A description of the study design, patient recruitment and data collection can be found in an earlier paper [20]. Written informed consent was obtained from all patients.
Primary isolation and drug susceptibility testing
Sputum specimens were shipped from Uzbekistan and Turkmenistan throughout the survey to the Supra-National Reference Laboratory (SRL) in Borstel, Germany. Primary isolation of mycobacterial isolates was performed as described elsewhere [22]. All isolates were identified as M. tuberculosis using gene probes (ACCUProbe, GenProbe, San Diego, USA), and standard biochemical procedures. Drug susceptibility testing (DST) was performed by using the proportion method on Löwenstein-Jensen medium and/or the modified proportion method in BACTEC 460TB (Becton Dickinson Microbiology Systems, Cockeysville, USA) according to the given instructions.
IS6110 DNA RFLP fingerprinting and spoligotyping analysis
Extraction of genomic DNA from the mycobacterial strains and DNA fingerprinting using IS6110 as a probe were performed according to a standardized protocol as described elsewhere [23]. Additionally, all isolates were analysed by the spoligotyping technique as described previously by Kamerbeek et al. [24]. The molecular typing data were analysed with the Bionumerics software (Windows NT, version 3.0; Applied Maths, Kortrijk, Belgium) as instructed by the manufacturer. The spoligotyping data were used to additionally confirm strain relationships and for the identification of Beijing genotype isolates (no hybridization to spacers 1–34, hybridization to spacers 35–43).
Statistical analyses
All clinical and laboratory data were entered into a database using Epi-info (6.04, CDC. Atlanta, GA, USA). Chi square analysis was used for comparisons of proportions. Logistic regression analysis was performed to identify variables independently associated with MDR-TB and Beijing genotype (SPSS version 10.0, SPSS Inc., Chicago, IL, USA).
Results
Out of the 416 strains included in the drug resistance survey, 397 were available for the IS6110 DNA fingerprint and spoligotyping analysis performed in this study (19 strains did not grow at the time the DNA fingerprinting was started). This subset is comprised of 208 patients from Karakalpakstan and 189 patients from Dashoguz. The characteristics of the patients included in this molecular investigation did not differ from the characteristics of the complete study population of the drug resistance survey (data not shown). In brief, the final sample consisted of 239 male (60%) and 158 female (40%) patients. The age of the patients ranged from 11 years to 77 years, with a mean of 34 years. Across both regions, 203 were new cases (51%) and 194 (49%) had received previous tuberculosis treatment.
Molecular typing results
In general, a high degree of diversity of IS6110 DNA fingerprint patterns was observed among the strains analysed. For 382 isolates, clear-cut IS6110 banding patterns were obtained, while, 15 strains (3.8%) showed mixed banding patterns demonstrating double infection with two M. tuberculosis strains (Fig. 1). These findings could be confirmed by the presence of mixed spoligotype patterns showing hybridization to the Beijing-typical spacers 35 to 43 and to spacer sequences not present in Beijing genotype strains (Fig. 1). In all cases of double infection identified, the patients were infected with a non-Beijing and a Beijing genotype strain, mainly with a weak Beijing genotype background pattern (Fig 1). All typing experiments were repeated for these strains to exclude the possibility of DNA carry over contamination. Since no clear IS6110 band definition is possible in mixed strain isolates, the patients with mixed infections were excluded from further investigations. Resistance to first-line drugs among the remaining 382 isolates, stratified by previous tuberculosis treatment and Beijing genotype, are shown in Table 1.
Figure 1 IS6110 DNA fingerprint and spoligotype patterns of the isolates obtained from five randomly chosen patients with double infections.
Table 1 Resistance to anti-tuberculosis drugs stratified by previous tuberculosis treatment and for patients infected with a Beijing or a non-Beijing strain.
New cases Previously treated Total Beijing Non-Beijing OR (Beijing) (95% CI)
Total cases 198 184 382 190 192
Resistance to:
Ethambutol 14 (7%) 47 (26%) 61 (16%) 49 (26%) 12 (6%) 5.2 (2.6–10.8)
Rifampicin 15 (8%) 54 (30%) 69 (18%) 51 (27%) 18 (9%) 3.6 (1.9–6.6)
Pyrazinamide 6 (3%) 24 (13%) 30 (8%) 22 (12%) 8 (4%) 3.0 (1.2–7.6)
Streptomycin 68 (3%) 111 (60%) 179 (47%) 114 (60%) 65 (34%) 2.9 (1.9–4.6)
Isoniazid 47 (24%) 108 (59%) 155 (41%) 99 (52%) 56 (29%) 2.6 (1.7–4.1)
MDR-TB 15 (8%) 53 (29%) 68 (18%) 51 (27%) 17 (9%) 3.8 (2.0–7.1)
To determine prominent genotypes and strains with identical IS6110 and spoligotype patterns, a dendrogram was calculated based on the similarity of the IS6110 RFLP patterns (Fig. 2). Among the 382 strains included in this analysis, 190 (49.8%) were of the Beijing genotype (Fig. 2). The majority of Beijing genotype isolates showed the typical spoligotype pattern (hybridization to all of spacers 35–43 and no hybridization to spacers 1–34) and "classical" Beijing genotype IS6110 RFLP patterns with a similarity of more than 70% (Fig. 2). Thus, these isolates formed a well-defined branch in the dendrogram. However, three isolates had further spacer deletions and did not hybridize to spacers 37 and 38, 41, 42 or 43, and 40 and 41, respectively. Furthermore, another four strains showed the typical Beijing genotype spoligotype pattern, but had IS6110 patterns not showing the characteristic Beijing genotype signature (Fig. 2).
Figure 2 IS6110 DNA fingerprint and spoligotype patterns of the 382 strains analysed. Banding patterns are ordered by similarity in a dendrogram.
Surprisingly, we also identified a number of isolates (n = 19, 5%) that belonged to the Delhi genotype, which has been found to be the dominant strain type in the Delhi region of India [25]. These strains showed highly similar IS6110 RFLP patterns thus also forming a certain branch in the dendrogram and had spoligotype patterns described to be typical for this genotype.
Based on identical RFLP and spoligotype patterns, 152 strains (40%) were grouped in 42 clusters ranging in size from two to 21 cluster members as follows: 26 clusters with 2 isolates, 4 with 3 isolates, 4 with 4 isolates, 3 with 5 isolates, 2 with 7 isolates, 1 with 8 isolates, 1 with 14 isolates, and 1 with 21 isolates. Although a third of these strains were in small clusters with just 2 isolates (n = 52 isolates), a remarkable number of strains were in the 2 largest clusters (n = 35), together representing 23% of all clustered isolates and 9% of all strains included in the cluster analysis. All larger clusters (n ≥ 4) were composed of Beijing genotype strains and, overall, 104 (68%) out of the 152 clustered isolates belong to the Beijing genotype. Overall, 55% of Beijing strains were clustered compared to 25% of non-Beijing strains (p < 0.01). There was no statistically significant difference in clustering between new and previously treated cases, between sexes or among age groups.
Characteristics associated with Beijing genotype strains
To define any association of drug resistance with Beijing genotype, we determined the percentage of Beijing genotype strains among different categories of drug resistance and by previous tuberculosis treatment status (Table 2). The association between Beijing genotype and levels of drug resistance was strikingly similar for both new and previously treated patients, with increasing proportions of Beijing type observed among categories of drug resistance (chi squared, p ≤ 0.001). While only 38% of the fully susceptible isolates belonged to the Beijing genotype, 75% of MDR-TB patients were infected with Beijing strains. The association between Beijing genotype and individual drug resistance, although evident for all first-line drugs, was not consistent (Table 1). There was a stronger association with ethambutol resistance, followed by rifampicin and pyrazinamide resistance, with MDR-TB closely mirroring the association with rifampicin resistance (Table 1). In contrast to the Beijing type, the smaller number of strains identified to be of the Delhi genotype was not associated with drug resistance when compared to strains not of the Beijing or Delhi families.
Table 2 Percentage of Beijing genotype isolates among different categories of drug resistance, by tuberculosis treatment status.
Total cases Beijing infection (%)
New cases Fully susceptible 124 48 (39%)
Resistant to one drug only 35 16 (46%)
Poly-resistant (not MDR-TB) 24 14 (58%)
MDR-TB 15 11 (73%)
Total 198 89 (45%)
Previously treated cases Fully susceptible 53 20 (38%)
Resistant to one drug only 37 17 (46%)
Poly-resistant (not MDR-TB) 41 24 (59%)
MDR-TB 53 40 (76%)
Total 184 101 (55%)
Previously, we have reported a logistic regression model identifying factors related to MDR-TB infection in this region [20]. Significant factors were previous tuberculosis treatment, residence in Karakalpakstan and female gender. When Beijing genotype is added to this model, it joins these factors as a significant independent predictor of MDR-TB (OR = 3.6 95%CI 1.9–6.8).
To investigate patient factors associated with Beijing genotype, univariate and multivariable analyses were performed with Beijing genotype infection as the dependent variable. Factors analysed were: previous tuberculosis treatment, belonging to a cluster, region (Karakalpakstan or Dashoguz), sex, previous imprisonment, reported contact with a tuberculosis case, alcohol use, accompanying illness, and age. Within univariate analysis, the only significant association observed was with clustering (OR = 3.6, Table 3). This result was confirmed in the multivariable analysis.
Table 3 Factors associated with Beijing genotype infection (univariate and multivariable analyses).
No. Beijing Univariate OR (95% CI) Multivariable OR (95% CI)
Previous TB treatment 184 101 1.5 (1.0–2.2) 1.3 (0.8–2.0)
Being in a cluster 152 104 3.6 (2.4–5.6) 3.5 (2.3–5.5)
Region (Karakalpakstan) 198 107 1.4 (1.0–2.2) 1.3 (0.8–1.9)
Female gender 156 75 0.9 (0.6–1.3) 1.1 (0.7–1.8)
Previous imprisonment 67 40 1.6 (1.0–2.8) 1.3 (0.7–2.5)
Close contact with a TB case 39 21 1.2 (0.6–2.3) 1.0 (0.5–2.0)
Alcohol use 30 20 2.1 (1.0–4.7) 2.0 (0.8–4.6)
Accompanying illness 48 19 0.6 (0.3–1.2) 0.6 (0.3–1.3)
Age over 30 202 102 1.0 (0.7–1.6) 1.1 (0.7–1.8)
Total 382 190
Discussion
This study demonstrates that the Beijing genotype is a major cause of tuberculosis in this high incidence region of Central Asia. A strong association with drug resistance has been documented, independent from previous tuberculosis treatment. In addition, an association of Beijing genotype with clustering suggests that these strains are being transmitted throughout the community, possibly mediating the transmission of drug-resistant tuberculosis in this setting.
Since its description in 1995, the Beijing genotype of M. tuberculosis has elicited increasing attention. The prevalence of the Beijing genotype shows strong geographical variation from below 10% to more than 90% of the strains analysed [16]. High rates of Beijing genotype strains have been reported from some regions in Eastern Europe, such as Estonia, Azerbaijan, and Russia (Samara oblast, Archangel oblast, north-western region), while in Western European countries the prevalence of Beijing genotype is low [8,9,16,26-32]. Although a high prevalence of Beijing genotype strains appears to be confirmed for some regions of the world, direct comparison of data is difficult due to variations in strain identification methodology and different survey inclusion criteria utilised. Many of the studies performed so far have addressed subpopulations such as prison inmates or have provided only limited information on inclusion criteria [28,29,32]. These studies indicate high rates of Beijing genotype strains among these sub-populations, but definitive conclusions regarding prevalence in the general population are more difficult.
This study presents results on the prevalence of Beijing genotype among a representative cross-sectional sample in Central Asia. Since 50% of patients were found to be infected with Beijing genotype isolates, the data obtained confirm the importance of the Beijing genotype in this region and place this region in Central Asia among the Beijing genotype "high incidence" regions. Infection with Beijing genotype was not associated with particular subgroups of patients such as previous prison inmates or with a specific region suggesting that tuberculosis due to Beijing genotype is not restricted to specific parts of the population, but affects the general community.
This is of importance, since the Beijing genotype was additionally found to be significantly associated with drug-resistance and might be a driving force for the spread and emergence of MDR-TB. It is well known, that strains of the Beijing genotype have been involved in outbreaks of drug-resistant tuberculosis such as in the USA and in Russia [18,19]. However, there was no consistent association with drug resistance among 12 studies evaluated in the review of Glynn and colleagues [16]. Only three studies found a strong association with resistance to particular drugs, while the other studies showed none or even a negative association.
In this investigation a strong association between the Beijing genotype and resistance to single drugs as well as with MDR-TB was observed. Beijing genotype was also associated with clustering when compared to non-Beijing strains, indicative of more recent transmission [33]. These data are in accordance with the results of the few recent studies so far performed in Eastern Europe: addressing tuberculosis in prison populations [28,32], and the general population in the Archangel and Samara Oblasts and north-western regions of Russia and Estonia [8,27,29]. It is therefore likely that the Beijing genotype is a major cause of tuberculosis and particularly MDR-TB in wide areas of the former Soviet Union.
From our data, we cannot tell whether the Beijing genotype has emerged recently in the Aral Sea region or whether its long-term presence has been revealed. However, data from the Archangel oblast indicate that the Beijing genotype has been introduced in this region recently and has been strongly disseminating during the last few years [8]. Additionally, the Beijing genotype has also been reported to infect a rising number of patients on Gran Canaria Island, Spain, adding to the conclusion that strains of the Beijing genotype might have a selective advantage compared to other M. tuberculosis strains and spread more rapidly [34]. Therefore, in the former Soviet Union, the introduction of Beijing strains may have coincided with the deterioration of socio-economic conditions and the tuberculosis control system, and together this combination may contribute to the current MDR-TB epidemic.
The presence of multiple infection in 4% of the samples in this study confirms previous findings reporting mixed infections among small numbers of patients [35,36]. The simultaneous infection of patients with a non-Beijing and a Beijing strain determined here, might point to specific properties of the Beijing genotype. In the majority of cases identified, the fingerprint pattern typical for the Beijing genotype was weak compared to the other pattern, indicating that an exogenous re-infection with a Beijing genotype strain might have occurred in these patients, which then out-competed the first isolate. A high rate of patients simultaneously infected with a Beijing and non-Beijing strain was also reported in a recent study in Cape Town, South Africa [37]. In this investigation, 19% of all patients were simultaneously infected with Beijing and non-Beijing strains. The possibility that resistant Beijing genotype strains can efficiently super-infect patients receiving regular tuberculosis treatment was further suggested by cases of exogenous re-infection with MDR-TB Beijing isolates described in previous studies [38,39]. These findings suggest that susceptible or resistant Beijing genotype strains can potentially re-infect tuberculosis patients and in the case of resistant strains even whilst they are receiving regular tuberculosis therapy. This would be expected to result in a selective advantage for Beijing strains and would lead to a higher Beijing prevalence among drug resistance isolates.
Future studies are needed to clarify the characteristics of this important genotype of M. tuberculosis. Possible variations in host-pathogen interactions among strains of various genotypes needs to be identified and the role of Beijing genotype infection as a possible risk factor for treatment failure and/or drug resistance development must urgently be addressed in longitudinal studies in the affected high incidence regions. In short, the effect of this genotype on tuberculosis control efforts need to be further investigated.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Helen Cox: Conception and design of the study, acquisition, analysis and interpretation of data, drafting and revising of the article, given final approval to this version to be published
Tanja Kubica: analysis and interpretation of data, drafting and revising of the article, given final approval to this version to be published
Daribay Doshetov: analysis and interpretation of data, drafting and revising of the article, given final approval to this version to be published
Yared Kebede: Conception and design of the study, drafting and revising of the article, given final approval to this version to be published
Sabine Rüsch-Gerdes: Conception and design of the study, interpretation of data, drafting and revising of the article, given final approval to this version to be published
Stefan Niemann: Conception and design of the study, acquisition, analysis and interpretation of data, drafting and revising of the article, given final approval to this version to be published
Acknowledgements
The authors like to thank B. Schlüter, I. Radzio, T. Ubben and P. Vock for excellent technical assistance and both national and expatriate staff of Médecins Sans Frontières in the field. Parts of this work were supported by the Robert-Koch-Institute, Berlin, Germany and the EU Concerted Action project QLK2-CT-2000-00630. The World Health Organization provided a small grant to support the initial drug-resistance survey. None of the authors have any potential conflict of interest with regard to the publication of this study
==== Refs
Corbett EL Watt CJ Walker N Maher D Williams BG Raviglione MC Dye C The growing burden of tuberculosis: global trends and interactions with the HIV epidemic Arch Intern Med 2003 163 9 1009 21 12742798 10.1001/archinte.163.9.1009
World Health Organization Global Tuberculosis Control: Surveillance, Planning, Financing. WHO Report 2005 2005 Geneva, Switzerland: WHO WHO/HTM/TB/2005.349
World Health Organization Anti-tuberculosis Drug Resistance in the World: Third Global Report 2004 Geneva: Switzerland: WHO
World Health Organization Press release: Drug-resistant tuberculosis levels ten times higher in Eastern Europe and Central Asia 2004 Geneva, Switzerland: WHO
World Health Organization Guidelines for establishing DOTS-Plus pilot projects for the management of multidrug-resistant tuberculosis (MDR-TB) 2000 Geneva, Switzerland: WHO WHO/CDS/TB/2000.279
Portaels F Rigouts L Bastian I Addressing multidrug-resistant tuberculosis in penitentiary hospitals and in the general population of the former Soviet Union Int J Tuberc Lung Dis 1999 3 582 588 10423220
Perelman MI Tuberculosis in Russia Int J Tuberc Lung Dis 2000 4 1097 103 11144450
Toungoussova OS Sandven P Mariandyshev AO Nizovtseva NI Bjune G Caugant DA Spread of drug-resistant Mycobacterium tuberculosis strains of the Beijing genotype in the Archangel Oblast, Russia J Clin Microbiol 2002 40 6 1930 37 12037045 10.1128/JCM.40.6.1930-1937.2002
Drobniewski F Balabanova Y Nikolayevsky V Ruddy M Kuznetzov S Zakharova S Melentyev A Fedorin I Drug-resistant tuberculosis, clinical virulence, and the dominance of the Beijing strain family in Russia Jama 2005 293 22 2726 31 15941801 10.1001/jama.293.22.2726
Kato-Maeda M Bifani PJ Kreiswirth BN Small PM The nature and consequence of genetic variability within Mycobacterium tuberculosis J Clin Invest 2001 107 5 533 37 11238552
Lopez B Aguilar D Orozco H Burger M Espitia C Ritacco V Barrera L Kremer K Hernandez-Pando R Huygen K van Soolingen D A marked difference in pathogenesis and immune response induced by different Mycobacterium tuberculosis genotypes Clin Exp Immunol 2003 133 1 30 37 12823275 10.1046/j.1365-2249.2003.02171.x
Manabe YC Dannenberg AM JrTyagi SK Hatem CL Yoder M Woolwine SC Zook BC Pitt ML Bishai WR Different strains of Mycobacterium tuberculosis cause various spectrums of disease in the rabbit model of tuberculosis Infect Immun 2003 71 10 6004 11 14500521 10.1128/IAI.71.10.6004-6011.2003
van Crevel R Nelwan RH de Lenne W Veeraragu Y van der Zanden AG Amin Z van der Meer JW van Soolingen D Mycobacterium tuberculosis Beijing genotype strains associated with febrile response to treatment Emerg Infect Dis 2001 7 5 880 83 11747703
van Soolingen D Qian L de Haas PE Douglas JT Traore H Portaels F Qing HZ Enkhsaikan D Nymadawa P van Embden JD Predominance of a single genotype of Mycobacterium tuberculosis in countries of east Asia J Clin Microbiol 1995 33 12 3234 38 8586708
Bifani PJ Mathema B Kurepina NE Kreiswirth BN Global dissemination of the Mycobacterium tuberculosis W-Beijing family strains Trends Microbiol 2002 10 1 45 52 11755085 10.1016/S0966-842X(01)02277-6
Glynn JR Whiteley J Bifani PJ Kremer K van Soolingen D Worldwide occurrence of Beijing/W strains of Mycobacterium tuberculosis: a systematic review Emerg Infect Dis 2002 8 8 843 49 12141971
Anh DD Borgdorff MW Van LN Lan NT van Gorkom T Kremer K van Soolingen D Mycobacterium tuberculosis Beijing genotype emerging in Vietnam Emerg Infect Dis 2000 6 3 302 05 10827122
Narvskaya O Otten T Limeschenko E Sapozhnikova N Graschenkova O Steklova L Nikonova A Filipenko ML Mokrousov I Vyshnevskiy B Nosocomial outbreak of multidrug-resistant tuberculosis caused by a strain of Mycobacterium tuberculosis W-Beijing family in St. Petersburg, Russia Eur J Clin Microbiol Infect Dis 2002 21 596 602 12226690 10.1007/s10096-002-0775-4
Moss AR Alland D Telzak E Hewlett D JrSharp V Chiliade P LaBombardi V Kabus D Hanna B Palumbo L Brudney K Weltman A Stoeckle K Chirgwin K Simberkoff M Moghazeh S Eisner W Lutfey M Kreiswirth B A city-wide outbreak of a multiple-drug-resistant strain of Mycobacterium tuberculosis in New York Int J Tuberc Lung Dis 1997 1 2 115 21 9441074
Cox HS Orozco JD Male R Rüch-Gerdes S Falzon D Small I Doshetov D Kebede Y Aziz M Multidrug-resistant tuberculosis in central Asia Emerg Infect Dis 2004 10 5 865 72 15200821
World Health Organization Guidelines for the surveillance of drug-resistant tuberculosis 1997 Geveva, Switzerland WHO/TB/96.216
Kent P Kubica G Public Health Mycobacteriology: A guide for the level III laboratory Atlanta, Ga: US Department of Health and Human Services, Centres for Disease Control 1985
van Embden JD Cave MD Crawford JT Dale JW Eisenach KD Gicquel B Hermans P Martin C McAdam R Shinnick TM Small P Strain identification of Mycobacterium tuberculosis by DNA fingerprinting: recommendations for a standardized methodology J Clin Microbiol 1993 31 406 409 8381814
Kamerbeek J Schouls L Kolk A van Agterveld M van Soolingen D Kuijper S Bunschoten A Molhuizen H Shaw R Goyal M van Embden J Simultaneous detection and strain differentiation of Mycobacterium tuberculosis for diagnosis and epidemiology J Clin Microbiol 1997 35 4 907 14 9157152
Bhanu NV van Soolingen D van Embden JD Dar L Pandey RM Seth P Predominace of a novel Mycobacterium tuberculosis genotype in the Delhi region of India Tuberculosis (Edinb) 2002 82 2–3 105 12 12356462 10.1054/tube.2002.0332
Lillebaek T Andersen AB Dirksen A Glynn JR Kremer K Mycobacterium tuberculosis Beijing genotype Emerg Infect Dis 2003 9 12 1553 57 14720395
Kruuner A Danilovitsh M Pehme L Laisaar T Hoffner SE Katila ML Tuberculosis as an occupational hazard for health care workers in Estonia Int J Tuberc Lung Dis 2001 5 2 170 76 11258511
Pfyffer GE Strassle A van Gorkum T Portaels F Rigouts L Mathieu C Mirzoyev F Traore H van Embden JD Multidrug-resistant tuberculosis in prison inmates, Azerbaijan Emerg Infect Dis 2001 7 5 855 61 11747699
Mokrousov I Filliol I Legrand E Sola C Otten T Vyshnevskaya E Limeschenko E Vyshnevskiy B Narvskaya O Rastogi N Molecular characterization of multiple-drug-resistant Mycobacterium tuberculosis isolates from northwestern Russia and analysis of rifampin resistance using RNA/RNA mismatch analysis as compared to the line probe assay and sequencing of the rpoB gene Res Microbiol 2002 153 4 213 19 12066892 10.1016/S0923-2508(02)01311-6
Diel R Schneider S Meywald-Walter K Ruf CM Rüsch-Gerdes S Niemann S Epidemiology of tuberculosis in Hamburg, Germany: long-term population-based analysis applying classical and molecular epidemiological techniques J Clin Microbiol 2002 40 2 532 39 11825968 10.1128/JCM.40.2.532-539.2002
Borgdorff MW de Haas P Kremer K van Soolingen D Mycobacterium tuberculosis Beijing genotype, the Netherlands Emerg Infect Dis 2003 9 10 1310 13 14609469
Drobniewski F Balabanova Y Ruddy M Weldon L Jeltkova K Brown T Malomanova N Elizarova E Melentyey A Mutovkin E Zhakharova S Fedorin I Rifampin- and multidrug-resistant tuberculosis in Russian civilians and prison inmates: dominance of the Beijing strain family Emerg Infect Dis 2002 8 11 1320 26 12453364
Murray M Nardell E Molecular epidemiology of tuberculosis: achievements and challenges to current knowledge Bull World Health Organ 2002 80 6 477 82 12132006
Caminero JA Pena MJ Campos-Herrero MI Rodriguez JC Garcia I Cabrera P Lafoz C Samper S Takiff H Afonso O Pavon JM Torres MJ van Soolingen D Enarson DA Martin C Epidemiological evidence of the spread of a Mycobacterium tuberculosis strain of the Beijing genotype on Gran Canaria Island Am J Respir Crit Care Med 2001 164 7 1165 70 11673204
Braden CR Morlock GP Woodley CL Johnson KR Colombel AC Cave MD Yang Z Valway SE Onorato IM Crawford JT Simultaneous infection with multiple strains of Mycobacterium tuberculosis Clin Infect Dis 2001 33 6 42 74 10.1086/322635
Das S Narayanan S Hari L Mohan NS Somasundaram S Selvakumar N Narayanan PR Simultaneous infection with multiple strains of Mycobacterium tuberculosis identified by restriction fragment length polymorphism analysis Int J Tuberc Lung Dis 2004 8 2 267 70 15139459
Warren RM Victor TC Streicher EM Richardson M Beyers N Gey van Pittius NC van Helden PD Patients with active tuberculosis often have different strains in the same sputum specimen Am J Respir Crit Care Med 2004 169 5 610 4 14701710 10.1164/rccm.200305-714OC
Niemann S Rüsch-Gerdes S Richter E IS6110 fingerprinting of drug-resistant Mycobacterium tuberculosis strains isolated in Germany during 1995 J Clin Microbiol 1997 35 3015 3020 9399486
Kubica T Rüsch-Gerdes S Niemann S The Beijing genotype is emerging among multidrug-resistant Mycobacterium tuberculosis strains from Germany Int J Tuberc Lung Dis 2004 8 9 1107 13 15455596
|
16277659
|
PMC1299328
|
CC BY
|
2021-01-04 16:36:25
|
no
|
Respir Res. 2005 Nov 8; 6(1):134
|
utf-8
|
Respir Res
| 2,005 |
10.1186/1465-9921-6-134
|
oa_comm
|
==== Front
PLoS MedPLoS MedpmedplosmedPLoS Medicine1549-12771549-1676Public Library of Science San Francisco, USA 1632397410.1371/journal.pmed.0030031PerspectivesNeurology/NeurosurgeryOncologyOncologyNeurologyRadiotherapyChemotherapyGlioblastoma Multiforme—Treating a Deadly Tumor with Both Strands of RNA PerspectiveWeil Robert J Robert J. Weil is at the Brain Tumor Institute, Cleveland Clinic Foundation, Cleveland, Ohio, United States of America. E-mail: [email protected]
Competing Interests: The author declares that no competing interests exist.
1 2006 6 12 2005 3 1 e31Copyright: © 2006 Robert J. Weil.2006This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.
EGF Receptor-Targeted Synthetic Double-Stranded RNA Eliminates Glioblastoma, Breast Cancer, and Adenocarcinoma Tumors in Mice
Weil discusses a promising new technique to deliver double-stranded RNA molecules to cancer cells, described in a new animal study in PLoS Medicine.
==== Body
Glioblastoma—The Clinical Challenge
Gliomas are the most common primary tumors of the brain, with an incidence of about 25,000 new cases per year in the United States [1]. At least half of all gliomas exhibit aggressive, malignant behavior. Glioblastoma multiforme (GBM), in particular, is clinically and pathologically malignant [1–3]. Patients with GBM have a poor prognosis, with a median survival of one year with aggressive therapy; fewer than 5% will survive five years [1,4,5]. In spite of its seemingly low incidence, mortality from GBM accounts for 3%–4% of all cancer deaths each year in the US [1].
The mainstays of treatment include surgical resection, radiation, and chemotherapy. Once adjuvant therapy is completed, gliomas generally recur at the surgical resection margin(s), and tend to be more aggressive than at initial presentation (Figure 1). At this stage in the course of disease, most therapy is palliative [1–5]. With the exception of a few early-stage clinical trials, current antiglioma therapies have not yet taken advantage of specific genetic abnormalities that lead to and sustain cancer. A new study by Alexander Levitzki and colleagues in this issue of PLoS Medicine presents promising preclinical results that appear to do just this, using a novel ligand-directed method to deliver double-stranded RNA molecules to cancer cells [6].
Figure 1 Representative Clinical, Pathological, and Molecular Genetic Features of Glioblastoma Multiforme
The top panel shows an illustrative set of neuroimages from a patient with glioblastoma multiforme. On the left and in the center are gadolinium-enhanced T1-weighted axial magnetic resonance images from the day before and the day after resection of a large left frontal GBM. The patient was treated with postoperative radiation and chemotherapy. He presented again, 11 months after the first surgery, with stupor and a contrast-enhanced computed tomographic scan (right), which showed a massive and fatal recurrence.
The middle panel shows histopathological examples of this patient's tumor. In the left image, there is evidence of hypercellularity, pseudopalisading necrosis (small arrow), and vascular proliferation with hemorrhage (large arrow). In the center image, there is hypercellularity with intermittent mitotic figures (small arrows), while in the image to the right, there are several areas with florid endothelial proliferation (small arrows). See also Table 1. All figures are 200× magnification.
The bottom panel shows a schematic of current models of astrocytoma development and progression. The de novo pathway is located on top, and the secondary pathway is located on bottom. The principal genetic changes are noted for each pathway. Neuronal tumors and oligodendrogliomas (left top and bottom, respectively) appear to arise independently. Average survivals are noted for each astrocytoma type. While basic genetic features have been elaborated, they have been inconsistently found in GBMs, and as yet appear not to be consistently effective for targeted therapy. There remains a great deal unknown about the process by which these tumors progress to the most malignant state, whether the tumor is a de novo GBM or arises secondarily. Unknown steps from potential progenitor or pluripotent tumor cell(s) are indicated by small arrows.
Table 1 Histological Features and Prognosis in Patients with Glioma
+, presence of characteristic; −, absence of characteristic. Greater numbers of + signs indicate greater prevalence of the characteristic, whereas combinations of + and − indicate that the characteristic may or may not be present.
Pathologic and Molecular Features
Gliomas are primary brain tumors that display pathological and ultrastructural features of glial cell differentiation. Primary brain tumors are classified on the basis of presumed line of neuroepithelial differentiation: astrocytic, oligodendroglial, and ependymal (Figure 1). Astrocytomas predominate, making up 80%–85% of all glial neoplasms, and will be the focus of this Perspective.
Grading is performed on a scale, from low to high, according to a tumor's histological features (Figure 1; Table 1). World Health Organization grade IV tumors, the GBMs, are aggressive, invasive, destructive malignancies, with increased mitotic activity, pronounced angiogenesis, necrosis, and proliferation rates two to five times higher than grade III tumors [2]. Roughly 50% of all GBMs are primary or de novo in origin, while the other half arises secondarily from lower-grade tumors [2], often after some years of latency [2]. Current models of gliomagenesis coincide with the two clinically recognized forms of GBM, de novo and progressive (Figure 1).
Most de novo GBMs do not have alterations in TP53; rather, nearly all carry EGF receptor (EGFR) gene amplifications, often combined with gene rearrangements that lead to a constitutively active, truncated receptor. By contrast, progression from a low-grade to a high-grade glioma often involves the serial accumulation of genetic alterations that inactivate tumor suppressor genes—such as TP53, p16, RB, PTEN—or activate oncogenes such as MDM2, CDK4 and CDK6 [2–4]. Functionally, gliomas seem to arise along two competing paths [3–5,7]. The first path is altered growth factor signaling—for example, activation of the EGFR-Ras-mitogen activated protein kinase, platelet-derived growth factor, or Akt pathways—which, both independently and through pathway crosstalk, lead to cell proliferation, cell cycle progression, and apoptosis inhibition (Figure 1). The second path is direct dys-regulation of cell cycle arrest, such as p16ink4a control of Rb or p14arf modulation of MDM2 and Tp53, among others.
Current Diagnosis and Prognosis
At present, beyond the positive predictive value of increasing malignancy, as defined histopathologically (Table 1), survival of patients with GBM is predicated on clinical variables, including the patient's age and condition (Karnofsky performance score) at diagnosis, tumor location and extent of surgical resection, and administration of adjuvant radiotherapy and/or chemotherapy [1–5]. With respect to each modality—surgery, radiation, or chemotherapy—the survival advantage for each remains modest, on the order of a few months, with an average overall survival from the time of initial diagnosis of about 12 months. Therefore, therapies that promote a meaningful survival advantage, while promoting and enhancing quality of life, are urgently needed.
Targeting Tumors with dsRNA
Because EGFR alterations are a common feature of many malignant tumors, including non-small-cell lung and colon cancers and malignant melanoma, among others, a variety of techniques have been designed to target the EGFR and its downstream agents, including antibodies, antisense RNAs, and a large number of small molecule inhibitors [7,8]. While many of these efforts have met with some success in other cancer types, none have had profound or lasting activity against GBM. Levitzki and colleagues have used a different strategy to target cells overexpressing EGFR: they use synthetic, double-stranded RNAs (dsRNAs), linked to EGF, to obtain selective and efficient killing of EGFR overexpressing malignant gliomas in vitro and in vivo in a mouse model [6].
dsRNA motifs are central to immune regulation, and dsRNA may play several roles in eukaryotic cells—blocking tolerance to tumor-associated self- and foreign antigens; activation of RNaseL and protein kinase R, which effect transcriptional and translational inhibition while simultaneously spurring interferon expression; induction of apoptosis; activation of small RNA-mediated interference; promotion of extracellular or paracrine effects through secretion of interferons and other cytokines (a “bystander effect’); and release of dsRNAs from infected cells, thereby activating antigen-presenting cells [9,10].
Given this panoply of potential effects, Shir et al. coupled dsRNA (a polyinosine-cytosine or poly IC construct) to EGF, and demonstrated that EGFR-targeted poly IC induced rapid and pronounced apoptosis of EGFR overexpressing cells, but not of cells expressing low EGFR, no EGFR, or mutated constitutively active EGFR, which cannot bind EGF. A variety of cytokines, including interferon-α, Gro-α, and interferon-induced protein-10/CXCL10—all of which have been shown to have antitumor or antiproliferative activity—were also expressed by the tumor cells. Importantly, these results were replicated in vivo, where dsRNA treatment led to survival of all animals with intracranial tumors for greater than 244 days. In addition, dsRNA treatment was equally applicable in vitro and in vivo for two other EGFR overexpressing cell lines, A431 (a cervical carcinoma) and MDA-MD-468 (a breast carcinoma), suggesting that this approach has potential for other tumor types that overexpress EGFR.
Clinical Implications
While other treatments have had encouraging in vitro and in vivo debuts in animals, they have failed when translated to malignant gliomas in humans. Only clinical data will show whether the approach described above will be successful in human patients. However, there is reason for cautious optimism: based on recent advances in delivery of macromolecules to the brain, specifically by convection-enhanced delivery, pioneered by Edward Oldfield at the National Institutes of Health, it appears that ligand-guided delivery of dsRNA may hold significant clinical promise [11,12]. Convection-enhanced delivery permits selective delivery of heterogenous macromolecules to targeted-diseased regions within the brain both safely and efficiently, while minimizing or eliminating toxicities to the healthy brain or outside the central nervous system [11,12]. And since the system of Shir et al. can link dsRNAs to essentially any molecule, ligand-guided delivery might be broadly applicable to any cancer, and, quite likely, to benign disorders as well, so long as an endocytosed receptor is substantially overexpressed compared to normal cells. It appears to me that this is one method that should be fast-tracked to the clinic.
Citation: Weil RJ (2006) Glioblastoma multiforme—Treating a deadly tumor with both strands of RNA. PLoS Med 3(1): e31.
Abbreviations
dsRNAdouble-stranded RNA
GBMglioblastoma multiforme
==== Refs
References
Berger MS Wilson CB The gliomas. 1st edition 1999 Philadelphia WB Saunders 796
Kleihues P Cavenee W Pathology and genetics of tumors of the nervous system 2000 Lyon IARC Press 314
Holland EC Gliomagenesis: Genetic alterations and mouse models Nat Rev Genet 2001 2 120 129 11253051
Maher EA Furnari FB Bachoo RM Rowitch DH Louis DN Malignant glioma: Genetics and biology of a grave matter Genes Dev 2001 15 1311 1333 11390353
Lacroix M Abi-Said D Fourney DR Gokaslan ZL Shi W A multivariate analysis of 416 patients with glioblastoma multiforme: Prognosis, extent of resection, and survival J Neurosurg 2001 95 190 198
Shir A Ogris M Wagner E Levitzki A EGF receptor targeted synthetic double-stranded RNA eliminates glioblastoma, breast cancer, and adenocarcinoma tumors in mice PLoS Med 2006 2 e6 10.1371/journal.pmed.0030006
Hanahan D Weinberg RA The hallmarks of cancer Cell 2000 100 57 70 10647931
Kapoor GS O'Rourke DM Receptor tyrosine kinase signaling in gliomagenesis: Pathobiology and therapeutic approaches Cancer Biol Ther 2003 2 330 342 14508101
von Herrath MG Bot A Immune reponsiveness, tolerance and dsRNA: Implications for traditional paradigms Trends Immunol 2003 24 289 292 12810099
Fire A RNA-triggered gene silencing Trends Genet 1999 15 358 363 10461204
Bobo RH Laske DW Akbasak A Morrison PF Dedrick RL Convection-enhanced delivery of macromolecules in the brain Proc Natl Acad Sci U S A 1994 91 2076 2080 8134351
Laske DW Youle RJ Oldfield EH Tumor regression with regional distribution of the targeted toxin TF-CRM107 in patients with malignant brain tumors Nat Med 1997 3 1362 1368 9396606
|
16323974
|
PMC1307534
|
CC BY
|
2021-01-05 11:13:40
|
no
|
PLoS Med. 2006 Jan 6; 3(1):e31
|
utf-8
|
PLoS Med
| 2,005 |
10.1371/journal.pmed.0030031
|
oa_comm
|
==== Front
BMC BiolBMC Biology1741-7007BioMed Central London 1741-7007-3-241628394210.1186/1741-7007-3-24Research ArticlePaucity of chimeric gene-transposable element transcripts in the Drosophila melanogaster genome Lipatov Mikhail [email protected] Kapa [email protected] Dmitri A [email protected] Casey M [email protected] Department of Biological Sciences, Stanford University, Stanford, CA 94305, USA2 Faculty of Life Sciences, University of Manchester, Manchester M13 9PT, UK2005 12 11 2005 3 24 24 6 7 2005 12 11 2005 Copyright © 2005 Lipatov et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Recent analysis of the human and mouse genomes has shown that a substantial proportion of protein coding genes and cis-regulatory elements contain transposable element (TE) sequences, implicating TE domestication as a mechanism for the origin of genetic novelty. To understand the general role of TE domestication in eukaryotic genome evolution, it is important to assess the acquisition of functional TE sequences by host genomes in a variety of different species, and to understand in greater depth the population dynamics of these mutational events.
Results
Using an in silico screen for host genes that contain TE sequences, we identified a set of 63 mature "chimeric" transcripts supported by expressed sequence tag (EST) evidence in the Drosophila melanogaster genome. We found a paucity of chimeric TEs relative to expectations derived from non-chimeric TEs, indicating that the majority (~80%) of TEs that generate chimeric transcripts are deleterious and are not observed in the genome sequence. Using a pooled-PCR strategy to assay the presence of gene-TE chimeras in wild strains, we found that over half of the observed chimeric TE insertions are restricted to the sequenced strain, and ~15% are found at high frequencies in North American D. melanogaster populations. Estimated population frequencies of chimeric TEs did not differ significantly from non-chimeric TEs, suggesting that the distribution of fitness effects for the observed subset of chimeric TEs is indistinguishable from the general set of TEs in the genome sequence.
Conclusion
In contrast to mammalian genomes, we found that fewer than 1% of Drosophila genes produce mRNAs that include bona fide TE sequences. This observation can be explained by the results of our population genomic analysis, which indicates that most potential chimeric TEs in D. melanogaster are deleterious but that a small proportion may contribute to the evolution of novel gene sequences such as nested or intercalated gene structures. Our results highlight the need to establish the fixity of putative cases of TE domestication identified using genome sequences in order to demonstrate their functional importance, and reveal that the contribution of TE domestication to genome evolution may vary drastically among animal taxa.
==== Body
Background
The origin of genetic novelty is of great interest in evolutionary biology. As mutation is the ultimate source of all genetic variation, understanding the mutational processes that lead to novel genomic features such as new genes, expression patterns or system interactions is paramount. The most commonly invoked mutational source of genetic novelty (after point substitution) is either segmental or whole genome duplication [1,2]. More recently, the role of duplicative transposition – the copying and pasting of particular DNA sequences from one part of genome to another – has been shown to play an important role in the evolution of new genes (e.g. [3]). Evidence from the human and mouse genomes indicates that, in addition to providing the source of the transpositional machinery, transposable elements (or TEs) [4] can also provide the template DNA for new genes or regulatory sequences [5-11]. However, to understand the general role of TE domestication in eukaryotic genome evolution, it is important to assess the acquisition of functional TE sequences by host genomes in a variety of different species, and to understand in greater depth the population dynamics of these mutational events.
Here we have investigated the incorporation of TEs into mature transcripts in the fruitfly Drosophila melanogaster, a species about which much is known in terms of the sequence and function of genic and intergenic regions. To do so, we searched for potentially domesticated "chimeric" transcripts (i.e. transcripts containing both TE and host gene sequences) backed by experimental support in the form of expressed sequence tag (EST) evidence (cp. [10,11]). The focus of this study is gene-TE associations contained within mRNA transcripts (i.e. within exons or untranslated regions, UTRs), so here we do not consider TEs that are either wholly contained in introns or located in the immediate vicinities of genes. An advantage of our approach is that the gene-TE chimeras identified are supported by experimental evidence rather than just by coordinate overlaps or mere proximity (cf. [12,13]), and thus enriches for a subset of TE insertions that may contribute to functional gene evolution in the host.
In addition, we have assessed the presence in wild populations of gene-TE chimeras identified using the genome sequence, to gain insight into the evolutionary forces acting on these mutations in nature. Using a pooled-PCR strategy, we estimated population frequencies for a sample of chimeric TE insertions in North American strains of D. melanogaster. By comparing population frequencies of chimeric TEs to those of non-chimeric TEs of the same family from similar genomic contexts, we evaluated whether chimeric TEs generally segregate either at unusually high frequencies (indicating the action of adaptive selection) or at unusually low frequencies (indicating the action of purifying selection). These results also revealed which of the gene-TE chimeras detected in the genome sequence are likely to be constitutive components of the D. melanogaster transcriptome.
By comparing our set of gene-TE chimeras to the entire set of annotated genes and TEs in the D. melanogaster Release 3 euchromatin, we show that a chimeric TE insertion has a much lower probability than a non-chimeric TE insertion of existing in the sequenced strain. This extreme paucity of chimeric TEs can be explained by the simple fact that TE insertions generating chimeric transcripts are likely to be strongly deleterious for the host. However, we find that the population frequencies of observed chimeric TEs are generally indistinguishable from similarly paired non-chimeric TE insertions, and we find that some chimeric TE insertions can be found at high frequency in North American populations. This pattern indicates that chimeric TE insertions observed in the genome sequence do not differ substantially from non-chimeric TEs in their selective effects, and that the D. melanogaster transcriptome permits a low-level flux of chimeric transcripts that may contribute to the formation of new gene sequences. Finally, we discuss the possibility that chimeric transcripts explain the curious phenomenon of regulated somatic expression of TE transcripts in the developing Drosophila embryo.
Results
Identification of chimeric gene-TE transcripts in the D. melanogaster genome
In order to study the functional integration of TE sequences into host genes, we identified TE insertions present in mature transcripts of the D. melanogaster euchromatic Release 3 genome sequence. We call such transcripts "chimeric" as each of them has one component from a host gene and one from a TE insertion. In addition to using the standard methods in the field for directly finding genes and TEs that share overlapping coordinates or querying annotated transcripts directly for TE sequences [8,10,11], we also sought evidence for chimeric transcripts using a novel three-step process based on expressed sequence tags (ESTs) (see Materials and Methods). This indirect method of identifying gene-TE chimeras was necessary to avoid annotation biases resulting from the fact that "coding exons were not annotated in sequences with homology to transposable elements" [14] in the D. melanogaster genome.
In total, we found 63 protein-coding genes that produce chimeric transcripts supported by EST evidence (Table 1; for more information [see Additional file 1]). These chimeric transcripts involve 63 different TE insertions, but the relationship is not simply one-to-one: in two cases, TE insertions (FBti0019107 and FBti0020178, Table 1) occur in overlapping 3'UTRs of convergently transcribed neighboring genes producing two separate chimeric transcripts each (see Figure 1A); and in one case, three TE insertions are found in a chimeric transcript for a single gene (CG32021) on the 4th chromosome. In addition, we found one noncoding transcript, the αγ-element [15], which is generated by two TE insertions within a larger nest of TEs situated between the Hsp70 Ba and Bbb genes. Our screen appears to have high sensitivity as evidenced by the fact that we identified four of the five exonic TE insertions previously reported in [12] (we found no supporting EST evidence for the fifth gene CG7900); the single exonic jockey insertion in the gene CG6191 reported in [16]; and the chimeric transcript generated by a Doc insertion into the gene CHKov1 (CG10618) reported in [17,18]. We did not identify the Bari-1 insertion in cyp12a4 recently reported in [19], which is supported by EST evidence, since the region of overlap (18 bp) does not pass our length threshold.
We note that six of the 65 chimeric TE insertions identified by BLAST-based methods do not have corresponding TEs in the Release 3.2 annotation. However, unannotated TEs of the correct family can be found in the genome sequence for these chimeric TE insertions (Table 1). This result indicates that an unknown proportion of real TE insertions has not been annotated in the Release 3 genome sequence (see below). To be able to analyze aspects of chimeric TEs in the context of the genome annotation, we excluded these six TE insertions from the "annotated set" of 59 TE insertions, although we do consider them to be bona fide members of the "total set" of 65 potential gene-TE chimeras in the D. melanogaster genome.
Properties of chimeric gene-TE transcripts
Most of the 63 genes generating the total set of chimeric transcripts are of unknown function, but we did identify chimeric transcripts in 23 characterized protein-coding genes including brown (bw), a gene that appears to be a hot-spot for natural TE insertions [20] and is known to carry a viable mutation (bw1) in the sequenced strain [14]. Our in silico screen also identified a chimeric TE insertion generated by the serine protease encoding gene Tequila that has recently been shown to impair the transcription of this gene, but with no apparent phenotypic consequences [21]. A general analysis of the molecular function and cellular localization of the total set of genes with chimeric transcripts, however, did not indicate a significant enrichment of any particular Gene Ontology (GO) category (data not shown).
Relative to other non-chimeric TEs inserted in transcribed regions (i.e. intronic TE insertions), the annotated set of TEs present in chimeric transcripts is significantly enriched for LTR insertions (Figure 2A). This observation largely accounts for the fact that the annotated set is also enriched in long TEs (Figure 2B), since LTR insertions tend to be longer than other classes of TE insertion in the genome [14]. Furthermore, chimeric TEs have a greater tendency to be present in high-recombination areas of the genome than non-chimeric, intronic TE insertions (Figure 2C). However, the overabundance of chimeric TEs in regions of high recombination is not caused simply by the fact that chimeric transcripts are preferentially formed by LTR insertions, since high-recombination TE insertions are over-represented among the chimeric non-LTR (i.e. LINE-like, TIR and FB) elements even more strongly than among the chimeric LTRs (data not shown).
TE sequences are found in UTRs in most of the chimeric transcripts they generate: 38 of the 63 TE insertions are found in 3'UTRs, 23 in the 5'UTRs and 4 in coding exons. We note that these numbers total more than 63 because two TE insertions (chimeras 47 and 61 [see Additional File 1]) fall into multiple categories. The higher incidence of TEs in UTRs and specifically in 3'UTRs parallels findings in the human and mouse genomes [10,11]. The increased prevalence of TE insertions in 3'UTRs may be attributed to the increased average length of 3'UTRs (442 bp) relative to 5'UTRs (265 bp) in Drosophila [22] (as has been suggested previously to explain such patterns in the human genome [10]), or to the lower density of functional signals in 3' regions relative to 5' regions of genes. This pattern does not appear to result from biases in the EST libraries, since over 10 times more 5' ESTs were analyzed than 3' ESTs [23].
Surprisingly, the genes involved in chimeric transcripts are not always those nearest to the sites of the corresponding TE insertions. Four chimeric transcripts skip one or more genes between the gene and TE components of the transcript (chimeras 12, 18, 23 and 50; Table 1, Figure 1B and 1C), thereby creating nested or intercalated gene arrangements. The process of gene- or exon-skipping in chimeric transcript formation suggests a novel mutational mechanism to explain the surprisingly large proportion of nested genes in the D. melanogaster genome (many of which bear no hallmark of retroposition) [22,24], as well as the evolution of complex intercalated gene structures that cannot arise via simple mechanisms of gene duplication.
Paucity of TEs in mature transcripts indicates that chimeric TE insertions are generally strongly deleterious
Of the 1,566 valid TEs in the Release 3.2 annotation of the D. melanogaster genome sequence, we estimate that 59 are chimeric TE insertions with some component co-transcribed in an exon, 414 are transcribed but entirely contained within spliced intronic sequences, and 1,093 are entirely contained within intergenic sequences not currently annotated as transcribed. A similar rank order pattern of TE abundance in different functional compartments has been observed in the Arabidopsis thaliana genome [25]. These numbers of TE insertions deviate significantly from their expected proportions based on the genome annotation of the 116.8 Mb Release 3 sequence (p < 1 × 10-15) (Table 2). This deviation from expectations is the result of two factors: there are fewer TEs in transcribed regions than in intergenic regions (p < 1 × 10-15) [14], and there is a further reduction in exonic regions relative to intronic regions (p < 1 × 10-15). The reduction in transcribed regions, however, is not solely caused by under-representation in exonic sequences, since the number of intronic TE insertions is reduced relative to the number in intergenic regions (p < 1 × 10-15). Together, these results indicate that there is a paucity of chimeric TE insertions in the genome, and that the causes of this paucity go above and beyond the effects of simply being transcribed.
To estimate the extent to which the number of exonic TE insertions is reduced while controlling for the effect of transcription per se on the distribution of TEs, we use the number of intronic TEs and the length of the intronic compartment of the genome to estimate the proportion of unobserved chimeric TE insertions. The total length of intronic regions in the D. melanogaster genome is approximately 37.7 Mb and the total length of exonic regions is 28.2 Mb [22,26]. If the selective pressures on exonic TEs were similar in magnitude to those on intronic TEs we would expect to find approximately 414*(28.2/37.7) = 310 TE insertions in the predicted exonic (coding plus untranslated) regions of the genome. The fact that we detect only 59 chimeric TEs out of an expected 310 (or 19%) indicates that a chimeric TE insertion is much more likely to be highly deleterious to the organism than a non-chimeric TE insertion that is spliced out of a mature transcript. These results are consistent with previous findings in the human genome, that the proportion of TE-derived sequence increases with increasing distance upstream from the start of transcription [10].
These calculations are based on a comparison of the annotated set of chimeric TE insertions relative to the total set of annotated TE insertions. As noted above, however, our results reveal that an unknown proportion of TEs in the Release 3 sequence were not annotated in [14]. If we assume that the frequency of unannotated TEs in intronic regions is proportional to that of the unannotated TE insertions in our sample (~10%), the expected number of TE insertions in exonic regions would increase to 310*1.10 = 341. Thus, using the total set under this proportionality assumption, the percentage of chimeric TE insertions detected relative to expectation is little changed (65 out of 341, 19%). To the extent that the number of unannotated TE insertions in introns is proportionally higher than in our sample, the percentage of observed chimeric TE insertions decreases even further, strengthening the claim for a paucity of chimeric TE insertions relative to expectation.
Observed chimeric TEs are not under unusual selective pressures
We estimated that ~80% of the TEs that have been inserted into mature genic transcripts are immediately purged from the genome by strong purifying selection, and therefore are not observed in the sequenced strain. What about the remaining ~20% of chimeric TE insertions that we do detect? We can envisage three scenarios to explain the existence of these chimeric TE insertions: 1) they are under strong purifying selection, like the TE insertions we do not observe; 2) they are adaptive, contributing useful sequences to the host genome; or 3) they are neither particularly deleterious nor particularly advantageous in comparison to the observed non-chimeric TE insertions in the genome.
In order to evaluate these possibilities, we surveyed the frequencies of chimeric TE insertions in wild D. melanogaster populations. The presence of each TE was tested in six pools of 8–12 North American strains (Table 1, rightmost column) using a PCR procedure custom-designed for each chimeric TE (see Methods for details). Pool frequencies were used to estimate confidence bounds on population frequencies using a maximum likelihood procedure (Table 3, [see Additional file 2]; see Methods for details).
We were able to generate population data for 48 of the 59 annotated chimeric TE insertions. Twenty-seven chimeric TE insertions were found only in the sequenced strain, seven were found in all six-strain pools and 14 had intermediate pool frequencies. These proportions of absent (56%) and polymorphic (44%) chimeric TEs are very similar to a combined, non-random sample of 92 non-chimeric TE insertions with previously reported population frequency data that map to annotated Release 3 TEs: absent (58%) and polymorphic (42%) [12,17,27-29]. The negative effects of intronic TE insertions on transcription do not strongly affect this non-chimeric sample, since similar proportions of absent and polymorphic TE insertions are observed in intronic (60% absent, 40% polymorphic; n = 30) and intergenic (56% absent, 44% polymorphic; n = 62) regions.
To determine whether the chimeric TE insertions are, on aggregate, subject to unusual selective constraints, we compared each of their pool frequencies to those of similar, non-chimeric TE insertions (Table 4). By "similar," we mean that these TE insertions came from the same family as their chimeric counterparts, that they had similar lengths, and were inserted in areas with similar recombination rates (see Methods for details). Since the selective constraint on a TE insertion is expected to increase with its length and the recombination rate of its genomic neighborhood [17,30], we tried to bracket each chimeric TE with a pair of similar non-chimeric family members: one with slightly higher, and one with slightly lower, length and recombination rate (columns 4 and 6 of Table 4, respectively). Our null hypothesis was that the chimeric TE insertions are neither particularly deleterious nor particularly advantageous in comparison with their non-chimeric counterparts. If this null hypothesis is true, we expect the pool frequencies of non-chimeric TE insertions in column 5 of Table 4 to be no higher, and the pool frequencies in column 7 to be no lower, than those of the chimeric TE insertions in column 3.
For the set of 48 TE insertions for which we have population data, we cannot reject the null hypothesis of no difference in pool frequencies between chimeric and non-chimeric TE insertions. Neither the Wilcoxon one-sided test nor the Kruskal-Wallis test reject the null hypothesis in favor of the alternative that pool frequencies of chimeric TEs are significantly higher than those of their counterparts with greater lengths and recombination rates (p = 0.38 and p = 0.75, respectively; tests performed on the n = 34 TEs in Table 4 that have the appropriate counterparts). This indicates that, in general, the fact that a TE insertion is chimeric does not increase the likelihood that it is at higher population frequency and is therefore potentially adaptive. Similarly, we find no evidence that chimeric TEs in general have pool frequencies lower than those with shorter lengths and lower recombination rates (p = 0.15 for the one-sided Wilcoxon rank sum test, p = 0.30 for the Kruskal-Wallis rank sum test; n = 46). Thus, the fact that an observed TE insertion is chimeric does not increase the likelihood that it is deleterious.
While we do not provide evidence for unusual selection pressures acting on chimeric TE insertions overall, we do find a few exceptions to this general rule when TE insertions are analyzed on an individual basis. As shown in Figure 3, by comparing pool frequencies of chimeric TEs to those of the two types of non-chimeric counterparts, we detect evidence for two exceptional chimeric TE insertions. One, a Doc insertion (FBti0019430), which creates a truncated version of the putative choline transferase gene CHKov1 (CG10618), has a significantly elevated population frequency (chimera 44, Figure 3A) and has been reported previously to be a putatively adaptive TE insertion [17,18]. The second, a pogo (FBti0019206) insertion into the fructose-bisphosphate encoding gene fbp, has a significantly decreased population frequency (chimera 21, Figure 3B) and is likely to be more deleterious than similar non-chimeric pogo insertions.
Discussion
We conducted a thorough search for TE insertions in the mature transcripts of genes in the sequenced D. melanogaster genome. To do so we used three different computational methods, including a novel, indirect EST-based approach (see Materials and Methods). As with all EST-based bioinformatics methods, this new approach to finding gene-TE chimeras is subject to biases in EST library composition. Such an approach was necessitated by annotation biases in the Drosophila genome that would have caused any direct analysis of annotated transcripts to underestimate the number of putative chimeric transcripts in the genome. Despite these conflicting biases, most of the 63 genes generating chimeric transcripts were identified by more than one method [see Additional file 1], although each method revealed unique chimeric TE insertions. Thus, multiple complementary approaches should be used in genome-wide studies of TE domestication to overcome both annotation and methodological biases.
Even using multiple methods for detecting chimeric transcripts, we estimate that only 0.46% of protein coding genes in Drosophila generate chimeric transcripts. Clearly the number of chimeric genes would be expected to increase somewhat with better annotation and/or increased EST coverage. Nevertheless, the number of chimeric transcripts in the Drosophila genome is likely to be more than an order of magnitude less than in the human and mouse genomes, where an estimated 27% and 18% of genes contain TE sequences [11]. These results together also suggest a rank order relationship between the proportion of chimeric genes and the amount of TE DNA in a genome (human, 46.36%; mouse, 38.55%; fly, 5.3%) [31-33]; however, further studies are needed to evaluate the strength and generality of this trend. Even a low number of gene-TE chimeras, such as presently observed in the D. melanogaster genome, may in the long-term contribute to the evolution of new transcripts and help explain unusual aspects of genomic organization structures such as nested or intercalated genes.
The low number of chimeric transcripts observed is not just the result of random effects of sparse TE insertion or the deleterious effects of TEs on transcription in the D. melanogaster genome. In fact, we found far fewer chimeric TE insertions in the genome than expected, relative to the number of non-chimeric TE insertions found in introns. This result indicates that the majority of TE insertions that occur in mature gene transcripts have a much higher probability of being deleterious than non-chimeric, intronic ones. The paucity of chimeric TE insertions in exons relative to introns demonstrates that the deleterious effects of chimeric TE insertions must exceed the cost of simply being transcribed, and probably results from improper translation or disruption of other functions of the mRNA such as localization or stability. Many of these unobserved events may contribute to the genome-wide load of deleterious mutations found in natural populations of D. melanogaster [34,35].
Population frequencies of the chimeric TE insertions observed in the genome sequence of the isogenized y; cn, bw, sp strain on the whole do not differ significantly from those of their non-chimeric counterparts. This does not imply that chimeric TE insertions found in the sequenced strain have no effects on fitness; rather that the distribution of their fitness effects is not substantially different from that of the non-chimeric TE insertions located elsewhere in the genome. At worst the observed chimeric TE insertions may be weakly deleterious and counter-selected, in contrast to the unobserved chimeric TE insertions, which are presumed to be strongly deleterious and purged rapidly from the population.
There is, however, some indirect evidence that chimeric TE insertions may in fact be less weakly deleterious on average than non-chimeric TE insertions. If TE insertions are weakly deleterious, we expect a skew towards genomic regions of lower recombination where natural selection is less effective due to increased linkage between alleles of opposing selective effects [36]. This effect can be observed in the distribution of non-chimeric, intronic TE insertions, but is not observed in the distribution of chimeric TE insertions (Figure 2C). Thus, a typical observed chimeric TE insertion may in fact have a smaller negative effect on fitness than a typical non-chimeric TE insertion. This conclusion is supported by a lack of detectable fitness effects in direct experimental challenges on flies carrying the chimeric TE insertion detected in the Tequila (graal) gene [21].
The one TE insertion we did identify as putatively adaptive (chimera 44; Figure 3A) was previously identified in a randomly chosen set of ~60 TEs [17,18]. We conclude that, in a search for adaptive TE insertions, selecting chimeric TE insertions is no better than selecting TEs from the Drosophila genome at random. This is perhaps not surprising, considering our finding that there is nothing unusual about the fitness effects of observed chimeric TE insertions. It is possible, however, that our inability to detect a significant difference in selection pressures resulted from the relatively small sample of both chimeric and control TE insertions studied here. Consideration of a larger number of strain pools will provide us with more statistical power and might show effects of chimerism on TE fitness that were not detected in this study.
Regardless of the forces that may have governed their history, we did identify seven chimeric TE insertions that appear to be at high frequency or possibly even fixed in North American populations of D. melanogaster. The existence of high frequency or fixed chimeric transcripts in the genome may provide a possible explanation for the curious observation of complex patterns of somatic gene expression exhibited by many LTR retrotransposons in D. melanogaster [37-40]. These largely-unexplored patterns of transcription are typically explained either by the existence of regulatory elements internal to the TE (internal enhancer model) or by the co-option of external cellular regulatory elements in the vicinity of a TE insertion (enhancer trap model) [39,41]. The presence of chimeric transcripts in the D. melanogaster genome demonstrated here suggests a third possible mechanism for the observed pattern of somatic TE expression: read-through transcription of a host gene into a TE and cross-hybridization to a TE specific probe. Under this model, regulated expression of a host gene that produces a chimeric transcript could be (mis)interpreted as regulated expression of the TE included in the chimeric transcript.
We sought evidence for the possibility of read-through transcription as an explanation for regulated TE expression by querying the second release of the BDGP in situ database [42,43] for embryonic expression patterns of the TEs and genes involved in chimeric transcripts detected in this study. Remarkably, as shown in Figure 4, we found that the embryonic expression pattern for developmental stages 11–16 of the gene CG12094 is almost identical to the expression pattern determined directly for the 412 element that is involved in the chimeric transcript generated by this gene.
Can read-through transcription from CG12094 explain the pattern of expression of the 412 element? We believe the answer to this question is no, for the simple reason that the probe used to determine the expression patterns of the 412 element (GM07634) shares no sequences for potential cross-hybridization with the chimeric CG12094 transcript (Figure 4). In addition, the TE insertion in CG12094 is not fixed, whereas the pattern of 412 element expression is similar among different strains (see [44]), suggesting that the presence of the 412 element insertion in CG12094 is not required for embryonic expression pattern of the 412 element. (In fact, these data taken together are more consistent with the stage 11–16 expression pattern of CG12094 detected by the RE52190 probe being generated by spurious cross-hybridization to transcripts emanating from 412 elements located elsewhere in the genome.) Thus in the case of the 412 element, we conclude that the best candidate gene in the D. melanogaster genome cannot explain somatic TE expression by production of a read-through chimeric transcript. Clearly more data will be necessary to evaluate the generality of this conclusion, but the lack of a role for read-though transcription in this case is generally consistent with the paucity and low population frequencies of the chimeric TE insertions in the D. melanogaster genome (Table 2) and with growing evidence for internal enhancer elements controlling regulated TE transcription [45-47].
Conclusion
In contrast to mammalian genomes, we found that fewer than 1% of Drosophila genes produce mRNAs that include bona fide TE sequences, and that the vast majority of potential chimeric TE insertions are likely to be deleterious and therefore unobserved in the genome sequence. Of those chimeric TE insertions that have weak enough negative fitness effects to have been observed in the sequenced D. melanogaster genome, over half are restricted to the sequenced strain and fewer than ~15% are likely to be fixed and therefore contribute to the origin of new gene sequences in the D. melanogaster genome. The relatively low numbers of fixed chimeric TE insertions also argue against read-through transcription as a predominant mechanism for generating patterns of somatic TE transcription in Drosophila embryos. These results also highlight the need to establish the fixity of putative cases of TE domestication identified in other genome sequences in order to demonstrate their functional importance, and indicate that the process of TE domestication may vary drastically among animal taxa.
Methods
in silico screen for chimeric gene-TE transcripts
Chimeric gene-TE transcripts were identified by three independent methods (with the following number codes used in Additional file 1): 1) a genomic coordinate intersection analysis; 2) a TE-to-gene BLAST analysis; and 3) a TE-to-EST-to-gene BLAST analysis. Coordinate overlaps were evaluated using the UCSC D. melanogaster table browser [48] and finding the intersection between the "FlyBase genes" and "FlyBase noncoding genes" tables, with a subsequent filter for those TE-gene overlaps >25 bp supported by EST evidence. For the TE-to-gene BLAST analysis, we sought chimeric TEs directly by querying each canonical TE sequence in version 7.1 of the BDGP TE data set [49] that had a representative in the Release 3.1 euchromatic genome annotation against the Release 3.1 annotated transcripts. For this analysis we used the combined output of hits from WU-BLASTN B = 10000 V = 10000 X = 3 M = 3-lcfilter-filter dust of >50 bp and >85% identity [50] together with NCBI-BLAST2 [51] hits of E < 1 × 10-10.
For the TE-to-EST-to-gene BLAST analysis, we developed a three-step process using WU-BLASTN with the following parameters: B = 10000 V = 10000 X = 3 M = 3-lcfilter-filter dust. First, each TE in the BDGP TE data set was used to query the BDGP EST database (ca. Dec 2002) containing 281,297 ESTs and complete cDNAs [23,52]. Second, ESTs with TE homology of >25 bp and >85% identity were aligned to the canonical TE sequence, and the non-TE component of the sequence was used to match the EST back to the corresponding host gene by querying transcripts in the Release 3.1 genome annotation [22]. Finally, the annotated host gene (±5000 bp) was used to query the TE database to ensure that a TE of the appropriate family is present in the genomic region, thereby filtering artifacts generated by EST library construction.
Transcripts from heterochromatic regions of the Release 3 genome were excluded from this analysis, as were genes labeled as "pseudogene" or unnamed genes with "existence uncertain" status in FlyBase. We also note that as in [14], we excluded from this analysis the enigmatic INE-1 element [53] that can be found in many transcripts [54], since this repetitive sequence is structurally distinct from all other TEs in the Drosophila genome.
Composition of DNA pools
A population of 64 individual strains from North America was combined into a total of 6 pools of 8 or 12 strains. The final concentration of each pool was 2.5 ng DNA of each individual strain per PCR reaction. The composition of each pool was as follows: Wi pool: Wi1, Wi3, Wi15, Wi18, Wi41, Wi45, Wi68, Wi77, Wi83, Wi98, Wi137, Wi148 – these strains were collected at the Wolfskill Orchard, Davis, CA and have been subjected to over 30 generations of brother-sister matings (gift and personal communication by Sergey Nuzhdin); We1 pool: We4, We7, We10, We11, We25, We44, We47, We50, We57, We60, WE67, We80; We2 pool: We13, We17, We21, We28, We33, We37, We63, WE70, We75, We83, We88, We91 – the strains in the two We pools were collected in Raleigh, NC and have been subjected to 10 – 15 generations of brother-sister matings (gift and personal communication by Greg Gibson); NA pool: Broward13, Broward5, Lake5, Okee14, Okee5, Orange1, Orange2, Paho4, Paho6, Paho9, Sebring12, Sebring17 – these isofemale strains were collected at various locations throughout North America (gift and personal communication by Jeff Birdsley); NB pool: NB1, NB6, NB7, NB8, NB12, NB13, NB14, NB16 – these isofemale strains were collected in New Buffalo, Michigan (gift and personal communication by Bettina Harr); CSW pool: 3B, 6D, 11D, 20C, 23D, 25C, 29B, 36D – these isofemale strains were collected at Countryside Winery, Blountville, Tennessee (gift and personal communication by Lev Yampolsky).
We note that some of the isofemale strains above may be heterozygous for a given TE insertion. This would lead to a slight increase in the effective number of strains in any given pool. However, such an increase is unlikely to have an effect on the qualitative nature of our results, as the addition of several strains to a pool generally has no significant effect on the confidence limits of the population frequency of a TE. For instance, in the section on population frequency estimation (below, also [see Additional file 2]), we show the extent to which the population frequency estimate remains the same when we treat 8-strain and 12-strain pools as if they were equivalent to each other.
PCR assays
The presence/absence of TEs in all strain pools was determined using the polymerase chain reaction (PCR). All PCR primers were designed using Primer 3 [55] and were checked with Virtual PCR [56]. All primers have a melting temperature of 63°C (+/-0.2°C) and were synthesized by Operon Biotechnologies, Inc. in 96 well plates. The primers are intended to assay for the presence of the TE insertion and consist of a "Left" primer that lies within the TE sequence and a "Right" primer that lies in the flanking region to the right of the TE insertion. Primer sequences used in this study can be found in Additional file 3. The presence of the TE insertion should produce a band of approximately 500 bp and the absence of the TE insertion should result in the absence of any band. On each plate there are 3 internal controls that should always produce a single band of predetermined size, designed to control for quality of PCR.
We also verified that the DNA concentrations were sufficient to detect the presence of TE in a single strain out of the 12 or 8 strains tested in the pool. Each plate of primers was assayed with a control pool comprising one of three North American pools (Wi, We1, or We2) with the addition of y; cn, bw, sp (sequenced strain) to control for primer design problems. The addition of y; cn, bw, sp should give a result indicating the presence of the TE insertion being assayed in all cases where primers were designed correctly. To be conservative, the concentration of the DNA from the y; cn, bw, sp strain was somewhat lower than that from the assayed strains. The PCR reaction mix was made using Redtaq Readymix from Sigma Aldrich (#R2523) and primers at a final concentration of 1 μmol/μl. The PCR conditions were: 94° for 5 s, 27 cycles of: 94° for 30 s, 62° for 30 s and 72° for 1 min. We note that for 83 TEs, the positive control PCR did not fail in any such cases, showing the presence of the TE; PCR with the same pool DNA lacking any TE showed its absence.
Estimation of TE population frequencies from pool frequencies
Given that a TE insertion is present in some of the North American strain pools and absent from others (i.e. given its pool frequency), we wished to calculate the likeliest frequency of this insertion in the entire North American population, as well as suitable confidence bounds around such a frequency estimate.
Let x1 (a number between 0 and 2) and x2 (a number between 0 and 4) be the respective numbers of 8-strain and 12-strain pools in which a particular element is present. Let y be the theoretical frequency of this element in the North American D. melanogaster population. The likelihood L, of any particular value of y given the observed values of x1 and x2 is proportional to the probability of obtaining such x1 and x2 if y has that value. That is,
L(y|x1, x2) ∝ Pr(x1 | y) × Pr(x2 | y) (1)
Where Pr(x1|y) is the probability that x1 out of two 8-strain pools contain the element and Pr(x2|y) is the probability that x2 out of four 12-strain pools contain the element, given that its overall frequency in the population is y.
The first term on the right hand side of equation (1) is equal to:
Where (1-y)8 is the probability that an element is not found in a given 8-strain pool, 1-(1-y)8 is the probability that it is, and the first term on the right hand side is the appropriate binomial coefficient. Similarly, the second term of equation (1) is equal to:
Substituting (2) and (3) into (1) and simplifying, we find that
Where k is an arbitrary multiplicative constant that absorbs the binomial coefficients in (2) and (3), since they are independent of the parameter y. In accordance with common practice, we make use of the log-likelihood function ln(L), which entails an arbitrary additive constant ln(k).
Additional file 2 provides three examples of the resulting log-likelihood functions. These functions correspond to the three possible combinations of x's that yield a total of four pools with detected element presence (i.e. for (x1, x2) equal to (0, 4), (1, 3) and (2, 2)). This file demonstrates that, given that the element is present in four out of six pools, estimation of population frequencies is relatively insensitive to the number of pools that contain eight or 12 strains. Therefore, to simplify the analysis, we combined all combinations of x1 and x2 under a common category such that x1 + x2 = 4.
For each log-likelihood function, the maximum likelihood estimate of the population frequency is the value of y at which the function reaches its maximum (middle column of Table 3). The confidence limits are determined by a likelihood ratio test of the values of y where the function drops below its maximus minus two (rightmost column of Table 3). The test statistic is the likelihood ratio of the 0-parameter model where y is fixed at the value of its maximum likelihood estimate to the 1-parameter model where y is allowed to vary. This statistic is distributed as a χ2 distribution with one degree of freedom. When the difference in log-likelihoods increases above 2, the likelihood ratio increases above e2 = 7.39, where e is the base of the natural logarithm. This value is the 99.3% quantile of the χ2 distribution (corresponding to p = 0.007, 1 d.f.). These confidence limits were used to set the error bars in Figures 3A and 3B. Note that in situations with more than one possible combination of x1 and x2 the two rightmost columns of Table 3 list values that are averaged over all possible combinations (see explanation for x1 + x2 = 4 above).
Estimation of genomic recombination rate in the neighborhood of each TE insertion
We estimated the recombination rate at each TE insertion site method using a method previously developed for the D. melanogaster genome [54]. This method combines the known physical and genetic distances between D. melanogaster genes to estimate the recombination rate profile of each chromosome as a second-degree polynomial function. An explanation of the method, and a tool that demonstrates its use, can be found on the world-wide web [57].
In Figure 2C, we classify chromosomal sites where the polynomial functions in [54] drop below zero as areas with "zero" recombination. We find that for the TE insertions in non-zero recombination areas, the median recombination rate is 2.75 cM / Mbp. Accordingly, we classify chromosomal sites with recombination rates above 0 and below 2.75 as areas with "low" recombination rates. The remaining chromosomal regions are labeled as areas of "high" recombination.
List of abbreviations
bp, base pairs; BDGP, Berkeley Drosophila Genome Project; BLAST, Basic Local Alignment Search Tool; EST, Expressed Sequence Tag; GO, Gene Ontology; LINE, Long Interspersed Nuclear Element; LTR, Long Terminal Repeat; Mbp, megabase pairs; PCR, Polymerase Chain Reaction; TE, Transposable Element; TIR, Terminal Inverted Repeat; UTR, Untranslated Region.
Authors' contributions
ML developed and carried out the population genomic analyses, and drafted the manuscript; KEL participated in the design of the population genetic study and gathered all molecular population genetic data; DAP helped conceive of the study, participated in the design and coordination of the population genetic and population genomic components of the study and helped draft the manuscript; CMB conceived of the study, conducted the bioinformatics analyses, participated in the analysis of the data and drafted the manuscript.
Supplementary Material
Additional file 1
Table of chimeric TE insertions in D. melanogaster Release 3 genome sequence with methods used for detection, location of TE in chimeric transcript, and supporting ESTs.
Click here for file
Additional file 2
Example of log-likelihood function for estimating population frequencies from pool frequencies (see methods for details).
Click here for file
Additional file 3
Table of PCR primers used in this study.
Click here for file
Acknowledgements
We thank Jeff Birdsley, Greg Gibson, Bettina Harr, Sergey Nuzhdin and Lev Yampolsky for the gifts of Drosophila strains and members of the DAP lab for helpful discussions. We thank Douda Bensasson and three anonymous reviewers for helpful comments on the manuscript. This work was funded by the Achievement Rewards for College Scientists Foundation through the Stanford Graduate Fellowship program (to ML); a NSF grant #0317171 (PI: DAP) and the Sloan and Hellman Fellowships (to DAP); a NIH training fellowship T32 HL07279 (PI: E. Rubin) and a USA Research Fellowship from the Royal Society (to CMB).
Figures and Tables
Figure 1 Examples of gene-TE chimeric transcripts in the D. melanogaster genome. Gene models are shown in blue, chimeric TE insertions are shown in red, and supporting EST clones are shown in black. A) A case of single TE insertion into the overlapping 3'UTRs of convergently transcribed genes creating two chimeric transcripts. B) A case of a chimeric transcript skipping several genes and creating a nested gene arrangement. C) A case of a chimeric transcript skipping an exon of a flanking gene creating an intercalated gene arrangement.
Figure 2 Properties of chimeric transcripts in the D. melanogaster genome. A) Proportions of 59 chimeric TEs in different element classes, compared to those of the 414 non-chimeric TEs found within intronic regions of genes. The proportion of LTR elements among the chimeric TEs is significantly greater than that of TEs found in introns. B) Proportions of 59 chimeric TEs in different length classes, compared to those of the 414 non-chimeric TEs within intronic regions of genes. The "low", "medium" and "high" length classes are defined according to the 33% and 66% length quantiles for the entire set of genomic TEs (748 and 3818 bp, respectively). The chimeric TEs show a significant enrichment for long TE insertions. C) Distribution of chimeric and non-chimeric TEs found within genes, partitioned by different recombination rates. Both distributions are compared against that of the number of genes found in each section of D. melanogaster euchromatin. See Methods for the definitions of "high", "low" and "zero" recombination. Note that the distribution of non-chimeric TEs deviates from the distribution of genes much more significantly than that of the chimeric TEs. In all three panels, the error bars on the numbers of chimeric insertions were obtained by assuming that the intronic proportion is the "true" probability p. Under a normal approximation, we expect the number of chimeric insertions to have mean np and variance np(1-p), where n is the number of chimeric elements. Based on this model, we constructed a 95% confidence interval around the observed number of chimeric elements that corresponds to the error bars in our figure. Error bars on the numbers of intronic insertions in panel C are based on the corresponding proportions of protein-coding genes.
Figure 3 Frequency of chimeric TE insertions compared with similar non-chimeric counterparts in North American strains of D. melanogaster. A) Pool frequencies of chimeric TEs (blue dots) versus those of their counterparts with lower length and recombination (red circles). Chimeric TE 44 has a significantly greater pool frequency than its counterpart, and was previously found to be adaptive [17, 18]. B) Pool frequencies of chimeric TEs (blue dots) versus those of longer counterparts with higher recombination rates (green triangles). Only one chimeric TE (number 21) has a significantly lower frequency than its counterpart. In both panels, the "population frequency" scale on the right-hand side gives maximum-likelihood estimates of the TE frequencies in the population (see Table 3 and the Methods for details).
Figure 4 mRNA in situ expression patterns of 412 element (A–D) and the chimeric 412-CG12094 transcript (E–H). Panel I shows a schematic of the 412 element and the in situ probes for the 412 element (GM07634) and CG12094 (RE52190). Shown are dorsal (A,C,E,G) and lateral (B,D,F,H) views of stage 10–11 (A,B,E,F) and stage 13 (C,D,G,H) embryos as extracted from the BDGP embryonic in situ database [42]. Note that the chimeric transcript from CG12094 has a nearly coincident pattern of expression with the 412 element at these stages of development.
Table 1 Chimeric TE insertions supported by EST evidence in the D. melanogaster Release 3 genome sequence. The leftmost column gives the gene(s) that generate(s) the chimeric transcript. FlyBase ID refers to the TE accession number in the Release 3.2 annotation. Rec. rate refers to the estimated recombination rate in the vicinity of the gene. Rightmost column gives the number of wild strain pools (out of six) where the TE insertion is present. An asterisk in the last column indicates that independent population frequency estimates are available for these TE insertions in [12, 17, 27-29].
Gene TE class TE family Chimera# FlyBase ID TE length (bp) Chrom. Arm Rec. rate Pool frequency
CG1098(Madm) TIR pogo 1 FBti0019306 185 3R 0 0
CG9766 LTR Burdock 2 FBti0019283 6412 3R 0 0
CG32306 LTR roo 3 FBti0020022 9097 3L 3.25 0
CG7110 LINE Doc 4 FBti0019166 4725 2L 2.83 0
CG3318(Dat) LTR 412 5 FBti0018872 7520 2R 3.16 0
CG4821(Tequila) LTR 412 6 FBti0020072 7502 3L 3.17 0
CG8166(unc-5) LTR Tirant 7 FBti0018948 8532 2R 3.44 0
CG6214 TIR pogo 8 FBti0019155 185 2L 3.07 0
CG1915(sls) LTR copia 9 FBti0020021 5145 3L 3.21 0
CG17632(bw) LTR 412 10 FBti0018870 7518 2R 3.31 0
CG2162 TIR S-element 11 FBti0020026 1733 3L 3.3 0
CG9181(Ptp61F) LINE Rt1a 12 FBti0020016 5192 3L 3.13 0
CG17274 LINE Doc 13 FBti0019412 4719 3R 3.09 0
CG32850 LTR Stalker2 14 FBti0020416 8118 4 0 0
CG7231 LTR 297 15 FBti0019135 6916 2L 4.02 0
CG7356 LTR copia 16 FBti0019136 5145 2L 4.01 0
CG7314(Bmcp) LINE Doc 17 FBti0020095 4709 3L 2.73 0*
CG11406 LTR 412 18 FBti0018873 7427 2R 3.15 0
CG3894 LINE G2 19 FBti0018918 1917 2R 2.99 0
CG5055(baz) LTR blastopia 20 FBti0019073 5031 X 2.88 0*
CG31692(fbp) TIR pogo 21 FBti0019206 186 2L 0 0
CG31177 LTR mdg1 22 FBti0019414 7338 3R 3.16 0
CG8776 LTR roo 23 FBti0019021 8313 2R 2.75 0
CG12885 LTR roo 24 FBti0020389 1051 3R 3.21 0*
CG32030 LINE I-element 25 FBti0020071 5348 3L 3.21 0
CG32594 LTR rover 26 FBti0019061 7469 X 3.61 0
CG17642(mRpL48) LTR mdg3 27a FBti0019107 267 2L 3.11 0*
CG17657 LTR mdg3 27b FBti0019107 267 2L 3.11 0*
CG7213 LINE Rt1a 28 FBti0020068 5177 3L 3.27 1
CG32684(alpha-Man-I) LTR roo 29 FBti0019615 9091 X 4.24 1
CG12094 LTR 412 30 FBti0019614 7441 X 4.23 1
CG18316 LTR 297 31 FBti0019977 6522 2R 0.68 1*
CG17697(fz) LINE X-element 32 FBti0020107 4728 3L 2.06 1*
CG18754 LTR roo 33 FBti0019420 427 3R 3.26 2
CG5130 LTR springer 34a FBti0020178 7542 3L 0 2
CG5976 LTR springer 34b FBti0020178 7542 3L 0 2
CG5656 TIR Tc1 35 FBti0020191 462 3L 0 4
CG31146 LTR invader3 36 FBti0020315 717 3R 0 4
CG14693 LTR 17.6 37 FBti0019354 7474 3R 1.03 4
CG11081(plexA) TIR Tc1 38 FBti0019510 530 4 0 5
CG32021 LTR invader4 39 FBti0019504 3082 4 0 5
CG18026(Caps) LINE F-element 40 FBti0020453 346 4 0 5
CG3812 TIR 1360 41 FBti0019634 376 X 4.07 5
CG32021 TIR 1360 42 FBti0019502 1075 4 0 6
CG18446 LTR roo 43 FBti0019985 427 2R 1.6 6*
CG10618 LINE Doc 44 FBti0019430 4512 3R 3.28 6
CG3136 LINE X-element 45 FBti0018950 1399 2R 0 6
CG32021 TIR transib3 46 FBti0019501 935 4 0 6
CG15347 TIR HB 47 FBti0019605 358 X 4.13 6
CG5541 TIR HB 48 FBti0019636 413 X 3.61 6
CG6191 LINE jockey 49 FBti0018988 265 2R 3.02 N.D.
CG10987 LTR roo 50 FBti0019665 427 X 0.94 N.D.
CG9527 LTR roo 51 FBti0019753 9098 2L 4.02 N.D.*
CG17521(Qm) TIR S2 52 FBti0020228 988 3L 0 N.D.
CR32865 LTR DM88 53 FBti0020348 168 3R 1.53 N.D.
CR32865 LTR invader1 54 FBti0020349 419 3R 1.53 N.D.
CG1090 LTR roo 55 FBti0019285 8325 3R 0 N.D.
CG1558(l(1)G0237) TIR pogo 56 FBti0019627 186 X 4.2 N.D.
CG1710 (Hcf) TIR 1360 57 FBti0020419 499 4 0 N.D.
CG1548(cathD) LTR Burdock 58 FBti0018882 6412 2R 0.52 N.D.*
CG3857 LTR opus 59 FBti0019540 7515 X 2.14 N.D.
CG32000 TIR 1360 60 N.A. 963 4 - N.D.
CG4494 (smt3) LTR mdg1 61 N.A. 51 2L - N.D.
CG6998 (ctp) LTR HMS-Beagle 62 N.A. 261 X - N.D.
CG3164 LTR McClintock 63 N.A. 80 2L - N.D.
CG7187(Ssdp) LTR HMS-Beagle 64 N.A. 212 3R - N.D.
CG9075(eIF-4a) LTR blood 65 N.A. 47 2L - N.D.
Table 2 Distribution of TEs by genomic compartment. Using the Release 3.2 annotation, the 116.8 Mbp D. melanogaster genome sequence was partitioned into exonic, intronic and intergenic DNA with exons taking precedence over introns, and introns over intergenic regions for genes with alternative splicing or promoter usage. χ2 values (degrees of freedom) are for tests of the number of TE insertions observed relative to expected proportions based on the total length of corresponding genomic compartment. P-values of all χ2 tests were <1 × 10-15.
Transcribed Exon Intron Intergenic Total χ2
Mbp 65.9 28.2 37.7 50.9 116.8 n.a.
observed # TEs 473 59 414 1093 1566 n.a.
expected (overall) - 377.86 505.20 682.94 1566 531.7 (2 df)
expected (transcribed vs. intergenic) 883.22 - - 682.78 1566 437.0 (1 df)
expected (exon vs. intron) - 202.44 270.56 - 473 177.7 (1 df)
expected (intron vs. intergenic) - - 640.48 866.53 1507 139.3 (1 df)
Table 3 Maximum likelihood (ML) estimates and bounds on TE insertion frequencies in the North American D. melanogaster population, given the number of pools that contain the TE insertion.
Number of pools containing TE (pool frequency) ML estimate of population frequency Likelihood bounds on population frequency
0 0.00% 0.00% – 3.0%
1 1.6% 0.09% – 7.2%
2 3.6% 0.58% – 11.1%
3 6.1% 1.5% – 15.8%
4 9.5% 2.9% – 22%
5 15% 5.1% – 35%
6 100% 11% – 100%
Table 4 For each chimeric TE (column 2), we give the number of strain pools in which the TE is present (column 3), the same for a similar TE with greater length in an area of higher recombination (columns 4 and 5), and for a similar TE with lower length inserted in an area with lower recombination (columns 6 and 7). For the first type of similar TE insertion, we expect slightly higher selective constraints, and thus slightly lower population frequency. The converse is true for the second type of similar TE insertion.
(1) # (2) chimeric TE (3) pool freq. (4) similar TE (expect lower pool freq.) (5) pool freq. (6) similar TE (expect higher pool freq.) (7) pool freq.
1 FBti0019306 0 FBti0019308 0 FBti0019323 0
2 FBti0019283 0 FBti0019236 3 FBti0019305 0
3 FBti0020022 0 FBti0019435 0 FBti0019025 0
4 FBti0019166 0 FBti0018904 0 FBti0019470 0
5 FBti0018872 0 FBti0018871 0 FBti0018874 0
6 FBti0020072 0 FBti0018871 0 FBti0018874 0
7 FBti0018948 0 FBti0018947 0
8 FBti0019155 0 FBti0020020 1 FBti0019565 1
9 FBti0020021 0 FBti0020031 0 FBti0019568 3
10 FBti0018870 0 FBti0018874 0
11 FBti0020026 0 FBti0020078 3
12 FBti0020016 0 FBti0019404 3
13 FBti0019412 0 FBti0019429 0 FBti0019470 0
14 FBti0020416 0 FBti0019960 5
15 FBti0019135 0 FBti0019611 0 FBti0019052 0
16 FBti0019136 0 FBti0019050 0
17 FBti0020095 0 FBti0018904 0 FBti0019377 2
18 FBti0018873 0 FBti0019380 0 FBti0018871 0
19 FBti0018918 0 FBti0019149 0 FBti0019171 4
20 FBti0019073 0 FBti0018961 0
21 FBti0019206 0 FBti0020172 5 FBti0019323 0
22 FBti0019414 0 FBti0019447 0 FBti0019989 3
23 FBti0019021 0 FBti0019556 0 FBti0019020 0
24 FBti0020389 0 FBti0019067 0 FBti0019024 0
25 FBti0020071 0 FBti0019057 0 FBti0019537 0
26 FBti0019061 0 FBti0019343 0
27 FBti0019107 0 FBti0019108 0 FBti0019971 5
28 FBti0020068 1 FBti0019404 3
29 FBti0019615 1 FBti0019608 0
30 FBti0019614 1 FBti0020033 0
31 FBti0019977 1 FBti0019353 4 FBti0019735 0
32 FBti0020107 1 FBti0019324 6
33 FBti0019420 2 FBti0019431 0 FBti0019438 0
34 FBti0020178 2 FBti0019466 0
35 FBti0020191 4 FBti0020429 6 FBti0020412 6
36 FBti0020315 4 FBti0020320 4 FBti0019329 6
37 FBti0019354 4 FBti0018861 0 FBti0019299 1
38 FBti0019510 5 FBti0019297 6 FBti0020429 6
39 FBti0019504 5 FBti0019234 6 FBti0019858 6
40 FBti0020453 5 FBti0019892 6 FBti0019492 6
41 FBti0019634 5 FBti0019613 4 FBti0019595 5
42 FBti0019502 6 FBti0019864 6 FBti0019518 6
43 FBti0019985 6 FBti0019363 0 FBti0019017 2
44 FBti0019430 6 FBti0020039 0 FBti0019417 0
45 FBti0018950 6 FBti0019783 6 FBti0019261 6
46 FBti0019501 6 FBti0019874 6
47 FBti0019605 6 FBti0020003 6
48 FBti0019636 6 FBti0020003 6
==== Refs
Lynch M Conery JS The evolutionary fate and consequences of duplicate genes Science 2000 290 1151 1155 11073452 10.1126/science.290.5494.1151
Ohno S Evolution by gene duplication 1970 London: George Allen and Unwin
Betran E Thornton K Long M Retroposed new genes out of the X in Drosophila Genome Res 2002 12 1854 1859 12466289 10.1101/gr.6049
McClintock B Controlling elements and the gene Cold Spring Harb Symp Quant Biol 1956 21 197 216 13433592
Brandt J Schrauth S Veith AM Froschauer A Haneke T Schultheis C Gessler M Leimeister C Volff JN Transposable elements as a source of genetic innovation: expression and evolution of a family of retrotransposon-derived neogenes in mammals Gene 2005 345 101 111 15716091 10.1016/j.gene.2004.11.022
Britten RJ Coding sequences of functioning human genes derived entirely from mobile element sequences Proc Natl Acad Sci U S A 2004 101 16825 16830 15546984 10.1073/pnas.0406985101
Sorek R Ast G Graur D Alu-containing exons are alternatively spliced Genome Res 2002 12 1060 1067 12097342 10.1101/gr.229302
Nekrutenko A Li WH Transposable elements are found in a large number of human protein-coding genes Trends Genet 2001 17 619 621 11672845 10.1016/S0168-9525(01)02445-3
Brosius J Genomes were forged by massive bombardments with retroelements and retrosequences Genetica 1999 107 209 238 10952214 10.1023/A:1004018519722
Jordan IK Rogozin IB Glazko GV Koonin EV Origin of a substantial fraction of human regulatory sequences from transposable elements Trends Genet 2003 19 68 72 12547512 10.1016/S0168-9525(02)00006-9
van de Lagemaat LN Landry JR Mager DL Medstrand P Transposable elements in mammals promote regulatory variation and diversification of genes with specialized functions Trends Genet 2003 19 530 536 14550626 10.1016/j.tig.2003.08.004
Franchini LF Ganko EW McDonald JF Retrotransposon-gene associations are widespread among D. melanogaster populations Mol Biol Evol 2004 21 1323 1331 15014149 10.1093/molbev/msh116
Ganko EW Bhattacharjee V Schliekelman P McDonald JF Evidence for the contribution of LTR retrotransposons to C. elegans gene evolution Mol Biol Evol 2003 20 1925 1931 12885961 10.1093/molbev/msg200
Kaminker JS Bergman CM Kronmiller B Carlson J Svirskas R Patel S Frise E Wheeler DA Lewis SE Rubin GM Ashburner M Celniker SE The transposable elements of the Drosophila melanogaster euchromatin: a genomics perspective Genome Biol 2002 3 RESEARCH0084 12537573 10.1186/gb-2002-3-12-research0084
Lis JT Prestidge L Hogness DS A novel arrangement of tandemly repeated genes at a major heat shock site in D. melanogaster Cell 1978 14 901 919 99245 10.1016/0092-8674(78)90345-8
Bartolome C Maside X Charlesworth B On the abundance and distribution of transposable elements in the genome of Drosophila melanogaster Mol Biol Evol 2002 19 926 937 12032249
Petrov DA Aminetzach YT Davis JC Bensasson D Hirsh AE Size matters: non-LTR retrotransposable elements and ectopic recombination in Drosophila Mol Biol Evol 2003 20 880 892 12716993 10.1093/molbev/msg102
Aminetzach YT Macpherson JM Petrov DA Pesticide resistance via transposition-mediated adaptive gene truncation in Drosophila Science 2005 309 764 767 16051794 10.1126/science.1112699
Marsano RM Caizzi R Moschetti R Junakovic N Evidence for a functional interaction between the Bari1 transposable element and the cytochrome P450 cyp12a4 gene in Drosophila melanogaster Gene 2005 16076534
Yang HP Tanikawa AY Kondrashov AS Molecular nature of 11 spontaneous de novo mutations in Drosophila melanogaster Genetics 2001 157 1285 1292 11238412
Munier AI Medzhitov R Janeway CA JrDoucet D Capovilla M Lagueux M graal : a Drosophila gene coding for several mosaic serine proteases Insect Biochem Mol Biol 2004 34 1025 1035 15475297 10.1016/j.ibmb.2003.09.009
Misra S Crosby MA Mungall CJ Matthews BB Campbell KS Hradecky P Huang Y Kaminker JS Millburn GH Prochnik SE Smith CD Tupy JL Whitfied EJ Bayraktaroglu L Berman BP Bettencourt BR Celniker SE de Grey AD Drysdale RA Harris NL Richter J Russo S Schroeder AJ Shu SQ Stapleton M Yamada C Ashburner M Gelbart WM Rubin GM Lewis SE Annotation of the Drosophila melanogaster euchromatic genome: a systematic review Genome Biol 2002 3 RESEARCH0083 12537572 10.1186/gb-2002-3-12-research0083
Stapleton M Carlson J Brokstein P Yu C Champe M George R Guarin H Kronmiller B Pacleb J Park S Wan K Rubin GM Celniker SE A Drosophila full-length cDNA resource Genome Biol 2002 3 RESEARCH0080 12537569 10.1186/gb-2002-3-12-research0080
Ashburner M Misra S Roote J Lewis SE Blazej R Davis T Doyle C Galle R George R Harris N Hartzell G Harvey D Hong L Houston K Hoskins R Johnson G Martin C Moshrefi A Palazzolo M Reese MG Spradling A Tsang G Wan K Whitelaw K Celniker S Rubin GM An exploration of the sequence of a 2.9-Mb region of the genome of Drosophila melanogaster: the Adh region Genetics 1999 153 179 219 10471707
Wright SI Agrawal N Bureau TE Effects of recombination rate and gene density on transposable element distributions in Arabidopsis thaliana Genome Res 2003 13 1897 1903 12902382
Celniker SE Wheeler DA Kronmiller B Carlson JW Halpern A Patel S Adams M Champe M Dugan SP Frise E Hodgson A George RA Hoskins RA Laverty T Muzny DM Nelson CR Pacleb JM Park S Pfeiffer BD Richards S Svirskas R Tabor PE Wan K Scherer SE Stapleton M Sutton GG Venter C Weinstock G Myers EW Gibbs RA Rubin GM Finishing a whole genome shotgun sequence assembly: release 3 of the Drosophila euchromatic genome sequence Genome Biology 2002 3 RESEARCH0079 12537568 10.1186/gb-2002-3-12-research0079
Maside X Bartolome C Charlesworth B S-element insertions are associated with the evolution of the Hsp70 genes in Drosophila melanogaster Curr Biol 2002 12 1686 12361573 10.1016/S0960-9822(02)01181-8
McCollum AM Ganko EW Barrass PA Rodriguez JM McDonald JF Evidence for the adaptive significance of an LTR retrotransposon sequence in a Drosophila heterochromatic gene BMC Evol Biol 2002 2 5 11914129 10.1186/1471-2148-2-5
Bartolome C Maside X The lack of recombination drives the fixation of transposable elements on the fourth chromosome of Drosophila melanogaster Genet Res 2004 83 91 100 15219154 10.1017/S0016672304006755
Nuzhdin SV Sure facts, speculations, and open questions about the evolution of transposable element copy number Genetica 1999 107 129 137 10952206 10.1023/A:1003957323876
Lander ES Linton LM Birren B Nusbaum C Zody MC Baldwin J Devon K Dewar K Doyle M FitzHugh W Funke R Gage D Harris K Heaford A Howland J Kann L Lehoczky J LeVine R McEwan P McKernan K Meldrim J Mesirov JP Miranda C Morris W Naylor J Raymond C Rosetti M Santos R Sheridan A Sougnez C Stange-Thomann N Stojanovic N Subramanian A Wyman D Rogers J Sulston J Ainscough R Beck S Bentley D Burton J Clee C Carter N Coulson A Deadman R Deloukas P Dunham A Dunham I Durbin R French L Grafham D Gregory S Hubbard T Humphray S Hunt A Jones M Lloyd C McMurray A Matthews L Mercer S Milne S Mullikin JC Mungall A Plumb R Ross M Shownkeen R Sims S Waterston RH Wilson RK Hillier LW McPherson JD Marra MA Mardis ER Fulton LA Chinwalla AT Pepin KH Gish WR Chissoe SL Wendl MC Delehaunty KD Miner TL Delehaunty A Kramer JB Cook LL Fulton RS Johnson DL Minx PJ Clifton SW Hawkins T Branscomb E Predki P Richardson P Wenning S Slezak T Doggett N Cheng JF Olsen A Lucas S Elkin C Uberbacher E Frazier M Gibbs RA Muzny DM Scherer SE Bouck JB Sodergren EJ Worley KC Rives CM Gorrell JH Metzker ML Naylor SL Kucherlapati RS Nelson DL Weinstock GM Sakaki Y Fujiyama A Hattori M Yada T Toyoda A Itoh T Kawagoe C Watanabe H Totoki Y Taylor T Weissenbach J Heilig R Saurin W Artiguenave F Brottier P Bruls T Pelletier E Robert C Wincker P Smith DR Doucette-Stamm L Rubenfield M Weinstock K Lee HM Dubois J Rosenthal A Platzer M Nyakatura G Taudien S Rump A Yang H Yu J Wang J Huang G Gu J Hood L Rowen L Madan A Qin S Davis RW Federspiel NA Abola AP Proctor MJ Myers RM Schmutz J Dickson M Grimwood J Cox DR Olson MV Kaul R Shimizu N Kawasaki K Minoshima S Evans GA Athanasiou M Schultz R Roe BA Chen F Pan H Ramser J Lehrach H Reinhardt R McCombie WR de la Bastide M Dedhia N Blocker H Hornischer K Nordsiek G Agarwala R Aravind L Bailey JA Bateman A Batzoglou S Birney E Bork P Brown DG Burge CB Cerutti L Chen HC Church D Clamp M Copley RR Doerks T Eddy SR Eichler EE Furey TS Galagan J Gilbert JG Harmon C Hayashizaki Y Haussler D Hermjakob H Hokamp K Jang W Johnson LS Jones TA Kasif S Kaspryzk A Kennedy S Kent WJ Kitts P Koonin EV Korf I Kulp D Lancet D Lowe TM McLysaght A Mikkelsen T Moran JV Mulder N Pollara VJ Ponting CP Schuler G Schultz J Slater G Smit AF Stupka E Szustakowski J Thierry-Mieg D Thierry-Mieg J Wagner L Wallis J Wheeler R Williams A Wolf YI Wolfe KH Yang SP Yeh RF Collins F Guyer MS Peterson J Felsenfeld A Wetterstrand KA Patrinos A Morgan MJ de Jong P Catanese JJ Osoegawa K Shizuya H Choi S Chen YJ Initial sequencing and analysis of the human genome Nature 2001 409 860 921 11237011 10.1038/35057062
Waterston RH Lindblad-Toh K Birney E Rogers J Abril JF Agarwal P Agarwala R Ainscough R Alexandersson M An P Antonarakis SE Attwood J Baertsch R Bailey J Barlow K Beck S Berry E Birren B Bloom T Bork P Botcherby M Bray N Brent MR Brown DG Brown SD Bult C Burton J Butler J Campbell RD Carninci P Cawley S Chiaromonte F Chinwalla AT Church DM Clamp M Clee C Collins FS Cook LL Copley RR Coulson A Couronne O Cuff J Curwen V Cutts T Daly M David R Davies J Delehaunty KD Deri J Dermitzakis ET Dewey C Dickens NJ Diekhans M Dodge S Dubchak I Dunn DM Eddy SR Elnitski L Emes RD Eswara P Eyras E Felsenfeld A Fewell GA Flicek P Foley K Frankel WN Fulton LA Fulton RS Furey TS Gage D Gibbs RA Glusman G Gnerre S Goldman N Goodstadt L Grafham D Graves TA Green ED Gregory S Guigo R Guyer M Hardison RC Haussler D Hayashizaki Y Hillier LW Hinrichs A Hlavina W Holzer T Hsu F Hua A Hubbard T Hunt A Jackson I Jaffe DB Johnson LS Jones M Jones TA Joy A Kamal M Karlsson EK Karolchik D Kasprzyk A Kawai J Keibler E Kells C Kent WJ Kirby A Kolbe DL Korf I Kucherlapati RS Kulbokas EJ Kulp D Landers T Leger JP Leonard S Letunic I Levine R Li J Li M Lloyd C Lucas S Ma B Maglott DR Mardis ER Matthews L Mauceli E Mayer JH McCarthy M McCombie WR McLaren S McLay K McPherson JD Meldrim J Meredith B Mesirov JP Miller W Miner TL Mongin E Montgomery KT Morgan M Mott R Mullikin JC Muzny DM Nash WE Nelson JO Nhan MN Nicol R Ning Z Nusbaum C O'Connor MJ Okazaki Y Oliver K Overton-Larty E Pachter L Parra G Pepin KH Peterson J Pevzner P Plumb R Pohl CS Poliakov A Ponce TC Ponting CP Potter S Quail M Reymond A Roe BA Roskin KM Rubin EM Rust AG Santos R Sapojnikov V Schultz B Schultz J Schwartz MS Schwartz S Scott C Seaman S Searle S Sharpe T Sheridan A Shownkeen R Sims S Singer JB Slater G Smit A Smith DR Spencer B Stabenau A Stange-Thomann N Sugnet C Suyama M Tesler G Thompson J Torrents D Trevaskis E Tromp J Ucla C Ureta-Vidal A Vinson JP Von Niederhausern AC Wade CM Wall M Weber RJ Weiss RB Wendl MC West AP Wetterstrand K Wheeler R Whelan S Wierzbowski J Willey D Williams S Wilson RK Winter E Worley KC Wyman D Yang S Yang SP Zdobnov EM Zody MC Lander ES Initial sequencing and comparative analysis of the mouse genome Nature 2002 420 520 562 12466850 10.1038/nature01262
Quesneville H Bergman CM Andrieu O Autard D Nouaud D Ashburner M Anxolabehere D Combined evidence annotation of transposable elements in genome sequences PLoS Comput Biol 2005 1 e22 10.1371/journal.pcbi.0010022
Mukai T Yamaguchi O The genetic structure of natural populations of Drosophila melanogaster. XI. Genetic variability in a local population Genetics 1974 76 339 366 4207116
Watanabe TK Yamaguchi O Mukai T The genetic variability of third chromosomes in a local population of Drosophila melanogaster Genetics 1976 82 63 82 814043
Hill WG Robertson A The effect of linkage on limits to artificial selection Genet Res 1966 8 269 294 5980116
Flavell AJ Ruby SW Toole JJ Roberts BE Rubin GM Translation and developmental regulation of RNA encoded by the eukaryotic transposable element copia Proc Natl Acad Sci U S A 1980 77 7107 7111 6261239
Parkhurst SM Corces VG Developmental expression of Drosophila melanogaster retrovirus-like transposable elements EMBO J 1987 6 419 424 3034591
Ding D Lipshitz HD Spatially regulated expression of retrovirus-like transposons during Drosophila melanogaster embryogenesis Genet Res 1994 64 167 181 7698641
Kearney JB Wheeler SR Estes P Parente B Crews ST Gene expression profiling of the developing Drosophila CNS midline cells Dev Biol 2004 275 473 492 15501232 10.1016/j.ydbio.2004.08.047
Arkhipova IR Lyubomirskaya NV Ilyin YV Drosophila Retrotransposons 1995 Austin, TX: R.G. Landes Co
BDGP Embryonic Expression Pattern Project
Tomancak P Beaton A Weiszmann R Kwan E Shu S Lewis SE Richards S Ashburner M Hartenstein V Celniker SE Rubin GM Systematic determination of patterns of gene expression during Drosophila embryogenesis Genome Biol 2002 3 RESEARCH0088 0088 12537577 10.1186/gb-2002-3-12-research0088
Brookman JJ Toosy AT Shashidhara LS White RA The 412 retrotransposon and the development of gonadal mesoderm in Drosophila Development 1992 116 1185 1192 1363543
Mozer BA Benzer S Ingrowth by photoreceptor axons induces transcription of a retrotransposon in the developing Drosophila brain Development 1994 120 1049 1058 8026319
Bronner G Taubert H Jackle H Mesoderm-specific B104 expression in the Drosophila embryo is mediated by internal cis-acting elements of the transposon Chromosoma 1995 103 669 675 7664613
Meignin C Dastugue B Vaury C Intercellular communication between germ line and somatic line is utilized to control the transcription of ZAM, an endogenous retrovirus from Drosophila melanogaster Nucleic Acids Res 2004 32 3799 3806 15263061 10.1093/nar/gkh708
UCSC D. melanogaster Genome Browser Gateway
BDGP Natural Transposable Element Project
Washington University BLAST Archives
NCBI BLAST
Stapleton M Liao G Brokstein P Hong L Carninci P Shiraki T Hayashizaki Y Champe M Pacleb J Wan K Yu C Carlson J George R Celniker S Rubin GM The Drosophila Gene Collection: Identification of putative full-length cDNAs for 70% of D. melanogaster genes Genome Res 2002 12 1294 1300 12176937 10.1101/gr.269102
Locke J Howard LT Aippersbach N Podemski L Hodgetts RB The characterization of DINE-1, a short, interspersed repetitive element present on chromosome and in the centric heterochromatin of Drosophila melanogaster Chromosoma 1999 108 356 366 10591995 10.1007/s004120050387
Singh ND Arndt PF Petrov DA Genomic heterogeneity of background substitutional patterns in Drosophila melanogaster Genetics 2005 169 709 722 15520267 10.1534/genetics.104.032250
Rozen S Skaletsky HJ Krawetz S, Misener S Primer3 on the WWW for general users and for biologist programmers Bioinformatics Methods and Protocols: Methods in Molecular Biology 2000 Totowa, NJ: Humana Press 365 386
Lexa M Horak J Brzobohaty B Virtual PCR Bioinformatics 2001 17 192 193 11238077 10.1093/bioinformatics/17.2.192
Drosophila melanogaster recombination rate calculator
|
16283942
|
PMC1308810
|
CC BY
|
2021-01-04 16:02:56
|
no
|
BMC Biol. 2005 Nov 12; 3:24
|
utf-8
|
BMC Biol
| 2,005 |
10.1186/1741-7007-3-24
|
oa_comm
|
==== Front
BMC Evol BiolBMC Evolutionary Biology1471-2148BioMed Central London 1471-2148-5-651628393810.1186/1471-2148-5-65Research ArticleTesting for adaptive evolution of the female reproductive protein ZPC in mammals, birds and fishes reveals problems with the M7-M8 likelihood ratio test Berlin Sofia [email protected] Nick GC [email protected] Department of Evolutionary Biology, Evolutionary Biology Centre, Uppsala University, Norbyvagen 18 D, 752 36 Uppsala, Sweden2 Current address: Department of Genetics and Genomics, Roslin Institute (Edinburgh), Roslin, Midlothian EH25 9PS, UK3 Department of Mathematics and Statistics, Lancaster University, Lancaster LA1 4YF, UK2005 10 11 2005 5 65 65 7 7 2005 10 11 2005 Copyright © 2005 Berlin and Smith; licensee BioMed Central Ltd.2005Berlin and Smith; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Adaptive evolution appears to be a common feature of reproductive proteins across a very wide range of organisms. A promising way of addressing the evolutionary forces responsible for this general phenomenon is to test for adaptive evolution in the same gene but among groups of species, which differ in their reproductive biology. One can then test evolutionary hypotheses by asking whether the variation in adaptive evolution is consistent with the variation in reproductive biology. We have attempted to apply this approach to the study of a female reproductive protein, zona pellucida C (ZPC), which has been previously shown by the use of likelihood ratio tests (LRTs) to be under positive selection in mammals.
Results
We tested for evidence of adaptive evolution of ZPC in 15 mammalian species, in 11 avian species and in six fish species using three different LRTs (M1a-M2a, M7-M8, and M8a-M8). The only significant findings of adaptive evolution came from the M7-M8 test in mammals and fishes. Since LRTs of adaptive evolution may yield false positives in some situations, we examined the properties of the LRTs by several different simulation methods. When we simulated data to test the robustness of the LRTs, we found that the pattern of evolution in ZPC generates an excess of false positives for the M7-M8 LRT but not for the M1a-M2a or M8a-M8 LRTs. This bias is strong enough to have generated the significant M7-M8 results for mammals and fishes.
Conclusion
We conclude that there is no strong evidence for adaptive evolution of ZPC in any of the vertebrate groups we studied, and that the M7-M8 LRT can be biased towards false inference of adaptive evolution by certain patterns of non-adaptive evolution.
==== Body
Background
Genes involved in reproduction and fertilization tend to evolve at faster rates than non-reproductive genes and it has been proposed that this rapid evolution is driven by positive Darwinian selection [1]. Mammalian sperm proteins have been extensively studied and they often show rapid divergence between closely related species [2-6]. The evolution of female reproductive proteins in vertebrate species has however received less attention, although evidence of positive selection has been provided for some female reproductive genes [7]. The precise nature of the selective pressures responsible for the accelerated evolution of reproductive genes are unknown, although plausible candidates include sexual selection, sexual conflict, and speciation reinforcement [1,8]. All three candidate evolutionary processes involve male-female interactions potentially manifested as sperm-egg interactions. So a way to test for these evolutionary processes is to compare patterns of evolution in genes involved in sperm-egg binding in (groups of) species, which differ with respect to sexual selection, sexual conflict or speciation reinforcement. For instance, in species with external fertilization, species recognition at the sperm-egg level should be more important than in species with internal fertilization, for which pre-fertilization mechanisms exist to avoid hybridisation. So if positive selection on reproductive genes is due to species recognition and hybridisation avoidance, then the signal of positive selection should be stronger in species with external fertilization than in species with internal fertilization. Furthermore, if species recognition at the sperm-egg level is driving the evolution of the genes involved in sperm-egg binding, then highly species-specific binding of sperm to the egg should yield a stronger signal of positive selection than less species-specific binding.
As an illustration of this general approach for determining the nature of positive selection in reproductive genes, we have investigated the molecular evolution of the gene encoding ZPC, the female reproductive glycoprotein zona pellucida C, in the three vertebrate groups of mammals, birds and fishes. The surface of the vertebrate oocyte is covered with an egg envelope, which is composed of several zona pellucida glycoproteins. ZPC is involved in the binding of sperm to the egg in mammals and birds, after which the acrosome reaction is triggered [9,10]. In general, when egg and sperm come from different mammalian species, binding of sperm to the egg envelope does not occur, thus the binding is species-specific [9]. From studies of sperm-egg binding in species from the avian order Galliformes (chicken and related species) it appears that the binding is not species-specific [11]. As a consequence, sperm-egg interactions in birds do not represent such a stringent species-specific barrier as they do in mammals. The fertilization mechanism in fish is rather different from that in mammals and birds. Most fish sperm lack an acrosome and penetrate the egg envelope via a discrete micropyle [12]. The function of the fish ZPC homologue is not entirely known, but studies in medaka (Oryzias latipes) suggest a function in guiding sperm to the micropyle, but it is unknown whether this process is species-specific.
In mammals, ZPC has previously been shown to be evolving by positive selection [7,13]. By considering the evolution of ZPC in birds and fishes as well as mammals we hoped to uncover variation in the levels of adaptive evolution among the three vertebrate groups. Since the three groups differ in their fundamental reproductive biology, any differences in adaptive evolution should be informative with regard to the underlying evolutionary mechanisms. By combining sequences from public databases and our own sequencing efforts, we constructed separate alignments of the ZPC coding sequence for 15 mammalian species, 11 avian species and six fish species. Our principle method for inferring adaptive evolution from ZPC sequence data was to apply LRTs of codon based models, which allow among-site variation in selection pressures. These models permit the inference of positive selection at a small proportion of sites, so that the signal of positive selection is not swamped by the (usually) much larger proportion of sites which are neutral or under negative selection. Such LRTs were used to demonstrate adaptive evolution of ZPC in mammals. However, there has been recent concern over high rates of false inference of positive selection with LRTs, particularly when applied to relatively small alignments with little sequence divergence [14-17]. Therefore we used three different LRTs and checked the evidence for positive selection by simulation studies.
Results
Overview of ZPC divergence
Using public available sequences and our own sequencing efforts, we produced three separate multiple species alignments of ZPC coding sequences for 15 mammals, 11 birds and six fishes. The total lengths of the alignments, which do not include start or stop codons but which do include gaps, are 1311 bp (437 codons) for mammals, 1335 bp (445 codons) for birds, and 1494 bp (498 codons) for fishes. Total tree length in substitutions per codon is 4.49 for mammals, 0.69 for birds, 4.49 for fishes. Thus there is relatively little sequence divergence in our sample of bird species compared to the other two groups of vertebrates.
We first analysed the ZPC interspecies alignments assuming the M0 model of no variation in ω among codons. ω is the nonsynonymous to synonymous rate ratio, also known as Ka/Ks or Dn/Ds, and is widely used as a measure of selective pressures on proteins assuming synonymous neutrality: ω<1 indicates negative selection, ω = 1 indicates neutrality, and ω>1 indicates positive selection. Average ω is 0.26 for mammals, 0.12 for birds and 0.32 for fishes. These low values indicate that the predominant mode of selection on the ZPC is purifying selection, and are consistent with the finding that domains within the ZPC gene are very well conserved across vertebrates [18]. Thus the M0 model of no variation in ω among codons gives no indication that ZPC is evolving unusually rapidly at the protein level.
Likelihood ratio tests of positive selection
LRTs of positive selection compare the fit of two nested models to the sequence data; a null model without adaptive evolution and an alternative model with adaptive evolution. Both models may invoke variation in ω among codons, but the null model is restricted to ω less than or equal to one, whereas the alternative model allows adaptive evolution with ω>1. If the alternative model provides a significantly better fit to the data then adaptive evolution is inferred. We considered three different LRTs, three different pairs of null and alternative models: M1a-M2a, M7-M8, M8a-M8 (see Methods).
Table 1 provides a summary of all the LRTs of positive selection performed on the multiple species alignments of ZPC coding sequences of mammals, birds and fishes. Let us first consider the evolution of ZPC in mammals, since M7-M8 LRTs have previously been used to demonstrate positive selection across a wide range of mammals [7] and more specifically in the Mus genus [13]. There is no indication of positive selection using the M1a-M2a test, since no sites are inferred to be under positive selection in the M2a model (hence p = 1). In contrast, the M8 model indicates that 9.7% of codons are subject to weak positive selection (ω = 1.32), with the chi-square approximation to the LRT (see Methods) indicating a significant improvement in fit from M7 to M8 (p = 0.004) and a suggestive improvement in fit from M8a to M8 (p = 0.084).
Table 1 Summary of PAML analyses of ZPC multispecies alignments. Significant results are in bold.
Mammals Fishes Birds
M0 ω = 0.26 ω = 0.32 ω = 0.12
M1a estimates P1 = 0.71 ω1 = 0.096 P1 = 0.69 w1 = 0.16 P1 = 0.93 ω1 = 0.07
P2 = 0.29 ω2 = 1.0 P2 = 0.31 w2 = 1.0 P2 = 0.07 ω2 = 1.0
M2a estimates P1 = 0.71 ω1 = 0.096 P1 = 0.70 w1 = 0.16 P1 = 0.95 ω1 = 0.07
P2+P3 = 0.29 ω2 = ω3 = 1.0 P2 = 0.0 ω2 = 1.0 P2 = 0.0 ω2 = 1.0
P3 = 0.30 ω3 = 1.04 P3 = 0.053 ω3 = 1.27
2Δlogl(M1a-M2a) 0 0.07 0.28
M1a-M2a p(X2) 1 0.97 0.87
M8a estimates P(beta) = 0.86 P(beta) = 0.72 P(beta) = 0.94
P1 = 0.16 ω1 = 1.0 P1 = 0.28 ω1 = 1.0 P1 = 0.065 ω1 = 1.0
M8 positive selection P = 0.097 ω = 1.32 P = 0.24 ω = 1.16 P = 0.049 ω = 1.29
2Δlogl(M7-M8) 10.9 14.9 3.67
M7-M8 p(X2) 0.004 0.0006 0.16
2Δlogl(M8a-M8) 3.00 0.75 0.32
M8a-M8 p(X2) 0.084 0.19 0.58
For birds, inferences of rare and weak positive selection are obtained for both the M2a (5.3% at ω = 1.27) and M8 (4.9% at ω = 1.29) models, although the chi-square approximations indicate no significant improvement in fits for alternative over null models (M1a-M2a p = 0.87, M7-M8 p = 0.16, M8a-M8 p = 0.58).
For fishes, the M2a model gives very little indication of positive selection (30% of codons with ω = 1.04), and the tiny increase in likelihood from M1a to M2a is not remotely significant (p = 0.97). The M8 model indicates that 24% of codons are subject to moderate positive selection (ω = 1.16). As with mammals the chi-square approximation indicates a significant improvement in fit for the M8 model over the M7 model (p = 0.0006). For the M8a-M8 comparison, the chi-square approximation indicates a non-significant improvement in fit (p = 0.19).
Simulation studies
The LRT analyses show that the three different LRTs give rather different views of positive selection in ZPC. The M1a-M2a test gives no indication of adaptive evolution in any of the three vertebrate groups. The situation is similar for the M8a-M8 test, although the result in mammals is at least suggestive. In contrast, the M7-M8 test gives highly significant evidence of adaptive evolution in mammals and fishes, and even though the M7-M8 LRT is not significant in birds the M7-M8 test gives the lowest p value of the three tests.
How can we explain the differences among the LRTs? There are a number of possible explanations: (1) there is no adaptive evolution in ZPC in vertebrates, and the chi-square approximation is biased to give false positives for the M7-M8 test (but not the other two LRTs); (2) there is no adaptive evolution in ZPC in vertebrates, and the M7-M8 test is biased to give false positives because it is not robust to the complex patterns of non-adaptive evolution in ZPC (but not the other two LRTs); and (3) there is adaptive evolution, and the M7-M8 test has greater power to reveal adaptive evolution than the other two tests. We performed simulations to examine these alternative explanations.
To test explanation (1) we performed parametric bootstrapping to check the chi-square approximation to the likelihood ratio statistic (see Materials and Methods). We performed parametric bootstrapping with 100 simulated data sets for the three LRTs in both mammals and fishes. In each case we found a good correspondence between the p value for the null model obtained by parametric bootstrapping and the p value obtained by the chi-square approximation (see Table 2). There seems to be no reason to suggest that any biases in the chi-square approximation could have generated the M7-M8 evidence of adaptive evolution, i.e. the p values obtained by parametric bootstrapping for the M7-M8 tests remain significant.
Table 2 Summary of simulation studies of ZPC LRTs.
p (X2) mammals p (parametric bootstrap) p (robustness) p (X2) fishes p (parametric bootstrap) p (robustness)
M1a-M2a 1 100/100 100/100 0.97 57/100 32/100
M7-M8 0.004 <1/100 25/100 0.0006 <1/100 52/100
M8a-M8 0.084 7/100 7/100 0.19 28/100 28/100
To test explanation (2) we performed simulations to check the robustness of the three LRTs to complex patterns of non-adaptive evolution as captured by the M8a model (see Materials and Methods). The p values for the robustness simulations in Table 2 indicate the probability of obtaining the observed LRT results where the data generated under the M8a model. If the LRTs are unbiased then this p value should be similar to the chi-square p value, but if a LRT is biased to give false positives then we would expect an increase from the chi-square p value to the robustness p value. We find strong evidence of a bias affecting the M7-M8 LRT: for both mammals and fishes the observed findings of positive selection could easily have been generated by the non-adaptive M8a model of codon evolution. In contrast the M8a-M8 test seems unbiased while the M1a-M2a test appears a little conservative.
To test explanation (3) we looked at the relative powers of the LRTs using simulations with a weakly selected positive selection class (ω distribution of 40% ω = 0, 40% ω = 0.25, 10% ω = 1.0 and 10% ω = 1.5). Power was measured as the proportion of the 100 data sets for which the LRT, as assessed by the chi-square approximation, gave a significant result with p < 0.05. The power of the M1a-M2a test was only 12%, much lower than the 54% power achieved by the M8a-M8 test, while the M7-M8 test has the greatest power of 88%.
Tests of positive selection in birds using polymorphism data
The fact that the total tree length of the avian multiple species alignment is relatively small means that methods using just inter-species divergence to infer positive selection are expected to lack power [14]. Therefore to study the evolution of avian ZPC in more detail we also performed various tests of positive selection making use of additional polymorphism data. The total number of segregating sites in the 46 chicken chromosomes was 56, of which 34 were intronic and 21 exonic, of which 3 were nonsynonymous and 18 synonymous. Tests based on allele frequency spectra failed to show significant deviations from neutrality or any evidence of recent selective sweeps. Tajima's D statistic was non-significantly positive (D = 0.074, p = 0.57), while the H test revealed a non-significant excess of high frequency alleles (H = -2.30, p = 0.13). The HKA test was performed to test for heterogeneity between the ratio intraspecific variation in chicken and interspecific divergence with the turkey outgroup in ZPC against a reference set of autosomal data from Sundstrom et al. [19].
The reference set contained only intronic sites, so only intronic sites in ZPC were considered. The HKA test did not show significant deviation from neutrality (p = 0.60). The nucleotide diversity in ZPC introns, θπ, is 8.9 × 10-3, higher by a factor of 1.37 than the average for several other intronic loci sequenced in the same individuals (θπ = 6.5 × 10-3) [19]. The relative increase in polymorphism is very nearly matched by the relative increase in divergence: the ZPC chicken-turkey intronic divergence of 0.129 is higher by a factor of 1.40 than the average of 0.092 from the study of Sundstrom et al. [19]. The McDonald-Kreitman test was performed to compare the nonsynonymous over synonymous ratio between divergence and polymorphism. For chicken polymorphism there were 3 nonsynonymous and 18 synonymous changes, while for divergence the numbers of nonsynonymous and synonymous changes down the chicken lineage were estimated by PAML to be 12 and 28. Thus there is a signal of positive selection, i.e. a relative excess of nonsynonymous divergence, but it is not significant (Fisher's exact test 1-T, p = 0.15).
Discussion
We have tested for signals of positive selection in the evolution of the ZPC gene in mammals, birds and fishes. The three different LRTs yielded some conflicting results (see Table 1), with the only two significant findings of positive selection both obtained using the M7-M8 test. The parametric bootstrapping simulations showed that this discrepancy between the LRTs was not due to biases in the chi-square approximation to the likelihood ratio statistic. The power simulations showed that the M7-M8 test is more powerful at detecting low levels of weak positive selection than the two other LRTs (as shown in [17]), but the robustness simulations showed that the M7-M8 test is biased towards false positives under patterns of non-adaptive evolution like those fitted by the M8a model to the ZPC gene. The bias in the M7-M8 test revealed by the robustness simulations was not apparent for the M1a-M2a and M8a-M8 tests, and this difference seems to provide the most likely explanation for our results: the highly significant M7-M8 results could have been generated by the bias in the test in the absence of adaptive evolution.
Thus we have no strong evidence of adaptive evolution in ZPC in vertebrates, with the best evidence being the suggestive p value of 0.08 for the M8a-M8 LRT on the mammalian sequence data. The use of avian polymorphism data in addition to interspecies data also failed to reveal significant evidence of positive selection.
Our results have some important methodological implications. The potential bias of the M7-M8 test has been suggested as a theoretical possibility by Swanson et al. [4], who pointed out that if the beta distribution on its own (the M7 model) fits the data poorly, then the M7-M8 test may generate a high proportion of significant tests even in the absence of positive selection. As far as we know, our robustness simulations provide the first demonstration of the M7-M8 bias for real data. The beta distribution is a natural distribution for modelling the variation in ω between 0 and 1 (e.g. see figures in the PAML manual), but a single beta distribution cannot provide a good fit when the real data have a bimodal distribution with two peaks, one located at ω = 1 and the other at an intermediate ω above 0 but less than 0.5. Thus the addition of an extra ω class in the M8 model is likely to give a large increase in likelihood, and if the position of the ω = 1 peak is overestimated at all then positive selection will be inferred.
We tested these ideas of why the M7-M8 LRT can lead to an excess of false positives by performing some additional robustness simulations under simple codon models. In all other respects except the ω distribution our set of 100 replicates was generated with the same parameters as the 5-taxon tree datasets detailed in Wong et al. [17]. The simulated ω distribution was 40% ω = 0, 40% ω = 0.25 and 20% ω = 1.0. In 18 out of 100 replicates the M7-M8 test gave a false positive result at the 5% level using the chi-square method, and six replicates gave a false positive result at the 1% level. So these robustness simulations confirm that the M7-M8 test can be biased under fairly simple scenarios. Since the scenario we have simulated seems biologically reasonable, i.e. a combination of some sites under very strong negative selection, some sites under moderate negative selection, and a small proportion of sites evolving neutrally, we believe that evidence of adaptive evolution obtained with the M7-M8 LRT alone should be treated with caution.
We also considered the performance of the M1a-M2a and M8a-M8 LRTs for the same simulated data, which led to an excess of false positives with the M7-M8 test. In none of the 100 replicates did the M1a-M2a test give a false positive result at the 5% level, indicating a conservative test. For the M8a-M8 test the levels of false positives were slightly elevated above expectations with nine false positives at the 5% level and two false positives at the 1% level, but these levels are reasonably close to null expectations. Thus the choice between the M1a-M2a and M8a-M8 LRTs can be seen as a tradeoff: the M8a-M8 test has greater power but also a greater risk of false positives.
We should emphasize the importance of distinguishing between problems with specific LRTs and the general approach using LRTs to test for adaptive evolution by comparing the likelihoods of different models of ω variation among codons. Our robustness simulations indicate that the M7-M8 LRT tends to produce false positives in the absence of adaptive evolution. Such a failure of a specific LRT does not mean that the general approach of LRTs is wrong, just that certain LRTs may fail under certain conditions. The more we know about how genes evolve the better we will be able to design models of among site variation in ω.
One final methodological point concerns the effects of changes in the size of the dataset. In a preliminary version of this study we considered a more limited dataset of just four fish species. For this dataset we obtained significant results of p = 0.044 for the M1a-M2a LRT and p = 0.032 for the M8a-M8 LRT. The addition of two more species (Carassius auratus and Pimephales promelas) increased the M1a-M2a p value to 0.97 and the M8a-M8 p value to 0.19. These two species only increased total length slightly from 4.14 to 4.49, so it is surprising that they had such a large effect on the M8a-M8 test. It seems likely that small datasets are likely to lead imprecise maximum likelihood estimates of parameters, particularly when fitting parameter rich codon models. Such concerns may be addressed by using Bayesian approaches to the inference of adaptive evolution, [20,21] which should account for parameter uncertainty.
Conclusion
So as far as biological conclusions go, we have no strong evidence of adaptive evolution in ZPC in vertebrates. This result obviously means that we cannot use comparative methods to analyse the selective causes of adaptive evolution in ZPC. Our study does however raise some interesting methodological issues, in particular we show that the patterns of evolution in the ZPC gene cause the M7-M8 LRT to be heavily biased towards false inference of adaptive evolution. Thus we urge caution in the use of the M7-M8 LRT, and would suggest the M1a-M2a and M8a-M8 LRTs as more reliable tests of adaptive evolution. Much of the evidence for pervasive adaptive evolution in reproductive proteins has been obtained using the M7-M8 LRT and our study raises the issue of whether these findings are genuine.
Methods
Sequencing of avian ZPC sequences
We have included eleven bird species from the order Galliformes in this study: black grouse (Tetrao tetrix), chicken (Gallus gallus), grey partridge (Perdix perdix), hazel grouse (Bonasa bonasia), pheasant (Phasianus colchicus), ptarmigan (Lagopus mutus), quail (Coturnix coturnix), red grouse (Lagopus lagopus), red-legged partridge (Alectoris rufa), sage grouse (Centrocercus urophasianus) and turkey (Meleagris gallopavo). In addition, we sequenced 23 male chickens. For information about these individuals see Sundstrom et al. [19]. The ZPC gene is located on chromosome 10 in chicken, according to the UCSC chicken genome browser. The gene has been fully sequenced in chicken [Genbank:AB031033], hence the exon-intron boundaries are known. We intended to sequence the entire gene, including exons and introns for all eleven species and the 23 male chickens. Despite many efforts, certain regions (nucleotides 1–373 and 715–773 in grey partridge, 315–372 and 411–445 in hazel grouse, 915–1007 in red-legged partridge and 414–442 and 790–815 in ptarmigan) could not be PCR amplified and were therefore not analyzed in these species.
We PCR amplified the gene from genomic DNA in five overlapping fragments using the following primer combinations: ZPC1051F/ZPCe2R, ZPCe2F/ZPCe5R, ZPCe4F/ZPCe6R, ZPC5F/ZPC7R and ZPC7F/ZPC9R (Primer sequences, primer positions and PCR conditions available on request). The same primers were used for all species except for ZPCe1Fmel, which was used together with ZPCe2R to amplify fragment 1 in turkey. The PCR reactions were performed in 20 μl volumes on Perkin Elmer 9600 Thermal Cyclers using 0,5 U AmpliTaq Gold (Applied Biosystems), 1.9–3 mM MgCl2 (Applied Biosystems) (exact concentrations avalailable on request), 0.08 mM dNTPs, 1x PCR Gold Buffer (Applied Biosystems), 5 pmol of each primer and 50 ng of template DNA. Turkey fragment 1 was amplified using Hotstar polymerase (Qiagen). PCR products were separated on 2% agarose gels, run in 0.5% TAE buffer, and visualized by ethidium bromide staining. PCR products were, prior to sequencing, purified with ExoSAP-IT reagent (Amersham Biosciences) followed by direct sequencing in forward and reverse directions using the DYEnamic™ ET DyeTerminator Kit (Amersham Biosciences). Sequencing primer sequences are available on request. Reactions were electrophoresed on a MegaBACE 1000 sequencing instrument (Amersham Bioscences). The sequences were edited in Autoassembler (Applied Biosystems) and overlapping forward and reverse sequences were compared to make consensus sequences. Complete diploid sequences from the 23 chicken individuals were assembled using ambiguity codes at heterozygote positions. Sequences have been deposited in public sequence databases [Genbank:AY628608-AY628630, Genbank:AY630568-AY630572, GenBank:DQ004565-DQ004569].
ZPC sequences of mammals and fishes
We collected a total of 15 mammalian ZPC sequences with accessions as follows: [Genbank:M20026] (mouse, Mus musculus), [Genbank:Y10823] (rat, Rattus rattus), [Genbank:X56777] (human, Homo sapiens), [Genbank:S71825] (marmoset), [Genbank:X82639] (macaque, Macaca radiata), [Genbank:D45070] (dog, Canis familiaris), [Genbank:D45068] (cat, Felis catus), [Genbank:D45065] (pig, Sus scrofa), [Genbank:NM_173974] (cow, Bos taurus), [Genbank:U05782] (rabbit, Oryctolagus cuniculus), [Genbank:AY598032] (fox, Vulpes vulpes), [Genbank:AY702973] (ferret, Mustela putorius), [Genbank:AF515621] (steppe lemming, Lagarus lagurus), [Genbank:AF304487] (Brandt's vole, Microtus brandti), [Genbank:M63629] (golden hamster, Mesocricetus auratus). The first eight sequences were taken from the analysis by Swanson et al. [7], while the remaining seven sequences were identified as homologs by BLAST searches [22], confirmed by annotation, and chosen for their effects on total branch length.
We collected six fish ZPC sequences with accessions as follows: [Genbank:NM_131331] (zebrafish, Danio rerio), [Genbank:L41638] (common carp, Cyprinus carpio), [Genbank:AF231708] (rainbow trout, Oncorhynchus mykiss), [Genbank:D38630] (Japanese medaka, Oryzias latipes), [Genbank:Z48974] (goldfish, Carassius auratus), and [Genbank:AF192407] (fathead minnow, Pimephales promelas). The first five sequences were confirmed as ZPC orthologs by Spargo and Hope [23], and the final sequence was identified as a homolog by a BLAST search and confirmed by annotation.
Construction of multiple species alignments
All alignments were generated using CLUSTALW [24]. The coding sequence alignments of DNA sequences were based on the protein alignments. Before we aligned the avian ZPC introns we removed regions of simple repeats, identified by the program Sputnik [25] since they make alignments unreliable and are a potential cause of bias due to elevated substitution and indel rates.
Likelihood ratio tests of positive selection
We used the codeml program in the PAML package [26] version 3.14 to perform likelihood ratio tests of positive selection. We considered models of codon evolution which allow for variation in ω among codons but assume the same distribution in all lineages. Likelihood ratio tests (LRTs) compare the maximum likelihoods of pairs of nested models, and when the models differ in whether they include codons under positive selection with ω>1, then the LRT is a test for positive selection [27,28]. There are many possible models of ω variation among codons, and hence many possible LRTs of positive selection. We performed three LRTs, which are thought to provide reliable tests of positive selection [4,17]. M1a-M2a LRT: The M1a model (one ω class between 0 and 1, and one class of ω = 1) is compared to the M2a model (same as M1a model plus an extra class of ω>1). M7-M8 LRT: The M7 model (a discretised beta distribution for ω between 0 and 1 with 10 equal class proportions) is compared to the M8 model (same as the M7 model plus an extra class of ω≥1). M8a-M8 LRT: The M8a model (same as M7 plus an extra class of ω = 1) is compared to the M8 model. For all LRTs, the first model is a simplified version of the second, with fewer parameters, and is thus expected to provide a poorer fit to the data (lower maximum likelihood). The first model is the null model without adaptive evolution and the second model is the alternative model with adaptive evolution, so a significant improvement in maximum likelihood supports positive selection.
The significance of likelihood ratio tests is usually calculated using the chi-square approximation, which states that at the asymptote when there is a large amount of data, then twice the difference in the log of maximum likelihood between the two models (the likelihood ratio statistic 2Δlogl) is distributed as a chi-square distribution with the degrees of freedom (df) given by the difference in the numbers of parameters in the two nested models. For the M1a-M2a comparison df = 2. For the M7-M8 and M8a-M8 comparisons the situation is more complicated due to problems with non-estimable parameters and parameter values being bounded. For the M7-M8 comparison the use of df = 2 is expected to be conservative, and for the M8a-M8 comparison the use of df = 1 is expected to be conservative.
For all LRTs, equilibrium codon frequencies were obtained using the average base composition at the three codon positions (CodonFreq = 2) and the transition-transversion rate ratio was estimated from the data. For the analysis of mammal and fish alignments of complete sequences, sites with ambiguity data were removed (cleandata = 1) since they correspond to indels. But for the analysis of the bird alignments, for which the sequences of some species were incompletely sequenced (see above), we did not remove sites with ambiguity data (cleandata = 0). One problem with the implementation of LRTs is the existence of suboptimal local maxima, which is overcome by use of several different starting points for the likelihood maximization. We checked carefully for local maxima in our analyses of real data, but not in our analyses of simulated data. We know that the PAML search algorithm did occasionally stall at suboptimal local optima in the analyses of our simulated data, since in around 1–2% of cases we obtained a higher likelihood with the simpler model, but such a low proportion is only expected to cause a slight downward bias to our simulation p values.
Simulation studies of LRTs
There are two potential problems with LRTs, which may generate false positives, i.e. the inference of adaptive evolution when there is none. (1) The first potential problem is that the chi-square approximation to the likelihood ratio statistic may be poor when data are limited. The problem in this case is not that the test itself is biased, but that the standard method for assessing significance of the likelihood ratio statistic does not hold. (2) The second potential problem is that certain LRTs may not be robust to certain patterns of non-adaptive evolution, i.e. they may have high rates of false positives. This case represents the more serious problem, that the LRT is inherently flawed in that it indicates positive selection even when there is non present. Both problems (1) and (2) were addressed by simulation studies performed using the evolver program in the PAML package.
We investigated the problem (1) by empirical parametric bootstrapping. For each LRT data was simulated using the maximum likelihood estimates of the parameters of the null model (M1a for M1a-M2a, M7 for M7-M8 and M8a for M8a-M8). The simulated data were then analysed using both the null and alternative models, and these results were used to generate the null distribution of the likelihood ratio statistic. Comparison of the likelihood ratio obtained from the real sequences with the null distribution of the likelihood ratios obtained from the simulated sequences indicates whether the null model of non-adaptive evolution is rejected in favour of the alternative model.
We addressed problem (2) by robustness simulations. We simulated data sets using the maximum likelihood parameter estimates obtained for the M8a model. Thus we simulated sequence data according to the richest model without positive selection, irrespective of which LRT was to be used to analyse the data. The simulated data were analysed using both the null and alternative models specific to each LRT, and these results were used to generate the null distribution of the likelihood ratio statistic. Comparison of the likelihood ratio obtained from the real sequences with the null distribution of the likelihood ratios obtained from the simulated sequences indicates whether the observed results could have been generated by non-adaptive evolution.
Phylogenetic trees
The PAML analyses require a single unrooted phylogenetic tree. The trees used in our analyses are shown in Figures 1 to 3. The mammalian tree is consistent with the emerging molecular view of mammalian phylogeny [29]. The fish tree is consistent with the tree in Spargo and Hope [23] for all species except Pimephales promelas, which was inferred to group with the other species in the Cyprinidae family using the stepwise addition method in PAML under the M0 model of no variation in ω among codons. Since Galliformes phylogeny is relatively uncertain we used MrBayes version 3.0b4 [30] to analyse the bird intron alignments. We specified a general time reversible model of nucleotide evolution, with a proportion of sites invariable and a gamma rate distribution for the remaining sites (using the MrBayes command "lset nst = 6 rates = invgamma"). The consensus bird tree we obtained differs slightly from the inferred phylogenies of Dimcheff et al. [31], but the differences are at nodes with relatively low bootstrap support in Dimcheff et al.'s maximum likelihood analysis of their mitochondrial sequences, and when we repeated our PAML analyses with a tree corresponding to the results of Dimcheff et al. we found no qualitative change to our results (data not shown).
Figure 1 The unrooted mammal tree used for the PAML analyses.
Figure 2 The unrooted fish tree used for the PAML analyses.
Figure 3 The unrooted bird tree used for the PAML analyses.
Tests of positive selection using polymorphism data
We used DnaSP 4.0 [32] to analyse patterns of polymorphism in our chicken ZPC sequences. McDonald-Kreitman tests [33] were performed with polymorphism in exons inferred using DnaSP and divergence estimated using PAML. The HKA test [34] was performed using the HKA computer program written by Jody Hey [35]. Intronic divergence was estimated using the baseml program in the PAML package under the Tamura-Nei model of nucleotide substitution [36]. We performed tests of selective sweeps using Tajima's D statistic [37] and the H statistic [38] using a computer program written by Justin Fay [39]. In order to infer the frequency of polymorphisms in chicken (required for the H test) we used parsimony with the turkey sequence as outgroup. Polymorphisms potentially due to CpG hypermutability were removed since parsimony inference may be poor in such cases. The probability of misinference was calculated as suggested by Fay and Wu [38] using the average intronic divergence between chicken and turkey (maximum likelihood distance of 0.129 given by PAML). The population scaled measure of recombination was estimated to be 65 across the whole gene using PHASE version 2.1 [40,41] to estimate a constant recombination rate with the unphased chicken data.
Authors' contributions
SB conceived the study and performed DNA sequencing of bird samples. NGCS performed the likelihood ratio test studies. Both authors designed the study, participated in sequence analyses of bird sequences, drafted the manuscript, and read and approved the final edition. Just after this manuscript was submitted Nick Smith tragically died in an accident. Nick was a splendid scientist and a great source of inspiration to me. He will be greatly missed.
Acknowledgements
We thank Cecilia Berg for DNA from Coturnix coturnix, Jacob Hoglund for DNA from Lagopus lagopus (JHGO 083–084), Perdix perdix (JHGO 127–128) and Tetrao tetrix, Thomas Ohlsson for DNA from Phasianus colchicus, Ettore Randi for DNA from Alectoris rufa, Swedish Museum of Natural History for DNA from Lagopus mutus (NRM 986102 and 986101) and Bonasa bonasia (NRM 996057 and 97/6584). We also thank Rasmus Nielsen and Ziheng Yang for discussions on the inference of positive selection. During the study NGCS was funded by Uppsala University, EPSRC and Lancaster University, while SB is funded by the Swedish Research Council.
==== Refs
Swanson WJ Vacquier VD The rapid evolution of reproductive proteins Nat Rev Genet 2002 3 137 144 11836507 10.1038/nrg733
Wyckoff GJ Wang W Wu CI Rapid evolution of male reproductive genes in the descent of man Nature 2000 403 304 309 10659848 10.1038/35002070
Torgerson DG Kulathinal RJ Singh RS Mammalian sperm proteins are rapidly evolving: evidence of positive selection in functionally diverse genes Mol Biol Evol 2002 19 1973 1980 12411606
Swanson WJ Nielsen R Yang Q Pervasive adaptive evolution in mammalian fertilization proteins Mol Biol Evol 2003 20 18 20 12519901
Glassey B Civetta A Positive selection at reproductive ADAM genes with potential intercellular binding activity Mol Biol Evol 2004 21 851 859 14963094 10.1093/molbev/msh080
Good JM Nachman MW Rates of protein evolution are positively correlated with developmental timing of expression during mouse spermatogenesis Mol Biol Evol 2005 22 1044 1052 15647515 10.1093/molbev/msi087
Swanson WJ Yang Z Wolfner MF Aquadro CF Positive Darwinian selection drives the evolution of several female reproductive proteins in mammals Proc Natl Acad Sci U S A 2001 98 2509 2514 11226269 10.1073/pnas.051605998
Swanson WJ Vacquier VD Reproductive protein evolution. Annu Rev Ecol Syst 2002 33 161 179 10.1146/annurev.ecolsys.33.010802.150439
Wassarman PM Jovine L Litscher ES A profile of fertilization in mammals Nat Cell Biol 2001 3 E59 64 11175768 10.1038/35055178
Bausek N Waclawek M Schneider WJ Wohlrab F The major chicken egg envelope protein ZP1 is different from ZPB and is synthesized in the liver J Biol Chem 2000 275 28866 28872 10979984 10.1074/jbc.275.37.28866
Stewart SG Bausek N Wohlrab F Schneider WJ Janet Horrocks A Wishart GJ Species specificity in avian sperm:perivitelline interaction Comp Biochem Physiol A Mol Integr Physiol 2004 137 657 663 15123173 10.1016/j.cbpb.2004.01.027
Iwamatsu T Yoshizaki N Shibata Y Changes in the chorion and sperm entry into the micropyle during fertilization in the teleostean fish, Oryzias latipes Dev Growth Differ 1997 39 33 41 9079033 10.1046/j.1440-169X.1997.00005.x
Jansa SA Lundrigan BL Tucker PK Tests for positive selection on immune and reproductive genes in closely related species of the murine genus mus J Mol Evol 2003 56 294 307 12612833 10.1007/s00239-002-2401-6
Anisimova M Bielawski JP Yang Z Accuracy and power of the likelihood ratio test in detecting adaptive molecular evolution Mol Biol Evol 2001 18 1585 1592 11470850
Suzuki Y Nei M Simulation study of the reliability and robustness of the statistical methods for detecting positive selection at single amino acid sites Mol Biol Evol 2002 19 1865 1869 12411595
Suzuki Y Nei M False-positive selection identified by ML-based methods: examples from the Sig1 gene of the diatom Thalassiosira weissflogii and the tax gene of a human T-cell lymphotropic virus Mol Biol Evol 2004 21 914 921 15014169 10.1093/molbev/msh098
Wong WS Yang Z Goldman N Nielsen R Accuracy and power of statistical methods for detecting adaptive evolution in protein coding sequences and for identifying positively selected sites Genetics 2004 168 1041 1051 15514074 10.1534/genetics.104.031153
Yang JC Hedrick JL cDNA cloning and sequence analysis of the Xenopus laevis egg envelope glycoprotein gp43 Dev Growth Differ 1997 39 457 467 9352200 10.1046/j.1440-169X.1997.t01-3-00007.x
Sundstrom H Webster MT Ellegren H Reduced variation on the chicken Z chromosome Genetics 2004 167 377 385 15166162 10.1534/genetics.167.1.377
Huelsenbeck JP Dyer KA Bayesian estimation of positively selected sites J Mol Evol 2004 58 661 672 15461423 10.1007/s00239-004-2588-9
Yang Z Wong WS Nielsen R Bayes empirical bayes inference of amino acid sites under positive selection Mol Biol Evol 2005 22 1107 1118 15689528 10.1093/molbev/msi097
Altschul SF Madden TL Schaffer AA Zhang J Zhang Z Miller W Lipman DJ Gapped BLAST and PSI-BLAST: a new generation of protein database search programs Nucleic Acids Res 1997 25 3389 3402 9254694 10.1093/nar/25.17.3389
Spargo SC Hope RM Evolution and nomenclature of the zona pellucida gene family Biol Reprod 2003 68 358 362 12533396 10.1095/biolreprod.102.008086
Thompson JD Higgins DG Gibson TJ CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice Nucleic Acids Res 1994 22 4673 4680 7984417
Sputnik program [http://espressosoftware.com/pages/sputnik.jsp]
Yang Z PAML: a program package for phylogenetic analysis by maximum likelihood Comput Appl Biosci 1997 13 555 556 9367129
Nielsen R Yang Z Likelihood models for detecting positively selected amino acid sites and applications to the HIV-1 envelope gene Genetics 1998 148 929 936 9539414
Yang Z Nielsen R Goldman N Pedersen AM Codon-substitution models for heterogeneous selection pressure at amino acid sites Genetics 2000 155 431 449 10790415
Springer MS Stanhope MJ Madsen O De Jong WW Molecules consolidate the placental mammal tree Trends in Ecology and Evolution 2004 19 430 438 16701301 10.1016/j.tree.2004.05.006
Ronquist F Huelsenbeck JP MrBayes 3: Bayesian phylogenetic inference under mixed models Bioinformatics 2003 19 1572 1574 12912839 10.1093/bioinformatics/btg180
Dimcheff DE Drovetski SV Mindell DP Phylogeny of Tetraoninae and other galliform birds using mitochondrial 12S and ND2 genes Mol Phylogenet Evol 2002 24 203 215 12144757 10.1016/S1055-7903(02)00230-0
Rozas J Rozas R DnaSP version 3: an integrated program for molecular population genetics and molecular evolution analysis Bioinformatics 1999 15 174 175 10089204 10.1093/bioinformatics/15.2.174
McDonald JH Kreitman M Adaptive protein evolution at the Adh locus in Drosophila Nature 1991 351 652 654 1904993 10.1038/351652a0
Hudson RR Kreitman M Aguade M A test of neutral molecular evolution based on nucleotide data Genetics 1987 116 153 159 3110004
HKA program [http://lifesci.rutgers.edu/~heylab/index.html]
Tamura K Nei M Estimation of the number of nucleotide substitutions in the control region of mitochondrial DNA in humans and chimpanzees Mol Biol Evol 1993 10 512 526 8336541
Tajima F Statistical method for testing the neutral mutation hypothesis by DNA polymorphism Genetics 1989 123 585 595 2513255
Fay JC Wu CI Hitchhiking under positive Darwinian selection Genetics 2000 155 1405 1413 10880498
H test program [http://www.genetics.wustl.edu/jflab/htest.html]
Li N Stephens M Modeling linkage disequilibrium and identifying recombination hotspots using single-nucleotide polymorphism data Genetics 2003 165 2213 2233 14704198
Crawford DC Bhangale T Li N Hellenthal G Rieder MJ Nickerson DA Stephens M Evidence for substantial fine-scale variation in recombination rates across the human genome Nat Genet 2004 36 700 706 15184900 10.1038/ng1376
|
16283938
|
PMC1308811
|
CC BY
|
2021-01-04 16:29:16
|
no
|
BMC Evol Biol. 2005 Nov 10; 5:65
|
utf-8
|
BMC Evol Biol
| 2,005 |
10.1186/1471-2148-5-65
|
oa_comm
|
==== Front
BMC GastroenterolBMC Gastroenterology1471-230XBioMed Central London 1471-230X-5-351628393510.1186/1471-230X-5-35Technical AdvanceMeasurement of the total antioxidant response using a novel automated method in subjects with nonalcoholic steatohepatitis Horoz Mehmet [email protected] Cengiz [email protected] Fusun F [email protected] Tevfik [email protected] Mehmet [email protected] Serpil [email protected] Necla [email protected] Ozcan [email protected] Harran University, Department of Internal Medicine, Sanliurfa, Turkey2 Harran University, Department of Gastroenterology, Sanliurfa, Turkey3 Harran University, Department of Endocrinology, Sanliurfa, Turkey4 Harran University, Department of Biochemistry, Sanliurfa, Turkey2005 11 11 2005 5 35 35 10 3 2005 11 11 2005 Copyright © 2005 Horoz et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Oxidative stress, an increase in oxidants and/or a decrease in antioxidant capacity, is one of the potential biochemical mechanisms involved in the pathogenesis of nonalcoholic steatohepatitis. We aimed to investigate the total antioxidant response using a novel automated method in nonalcoholic steatohepatitis subjects. As a reciprocal measure, we also aimed to determine total peroxide level in the same plasma samples.
Methods
Twenty-two subjects with biopsy proven nonalcoholic steatohepatitis and 22 healthy controls were enrolled. Total antioxidant response and total peroxide level measurements were done in all participants. The ratio percentage of total peroxide level to total antioxidant response was regarded as oxidative stress index.
Results
Total antioxidant response of subjects with nonalcoholic steatohepatitis was significantly lower than controls (p < 0.05), while mean total peroxide level and mean oxidative stress index were higher (all p < 0.05).
In subjects with nonalcoholic steatohepatitis, fibrosis score was significantly correlated with total peroxide level, total antioxidant response and oxidative stress index (p < 0.05, r = 0.607; p < 0.05, r = -0.506; p < 0.05, r = 0.728, respectively). However, no correlation was observed between necroimflamatory grade and those oxidative status parameters (all p > 0.05).
Conclusion
Nonalcoholic steatohepatitis is associated with increased oxidant capacity, especially in the presence of liver fibrosis. The novel automated assay is a reliable and easily applicable method for total plasma antioxidant response measurement in nonalcoholic steatohepatitis.
==== Body
Background
Non-alcoholic fatty liver disease (NAFLD) is increasingly recognized as a significant cause of morbidity in developed countries. NAFLD represents a spectrum of liver disease ranging from bland steatosis to severe steatohepatitis. The prevalence of non-alcoholic steatohepatitis (NASH) is 2.1–6.3% in the general population, rising to 9–40% in obese individuals with a body mass index (BMI) of 30 kg/m2 or more. Steatohepatitis can be progressive, cause fibrosis and cirrhosis, and can ultimately lead to liver failure and hepatocellular carcinoma in a minority of patients [1-4].
The pathogenesis of NASH remains unclear and the factors, which cause the progression from bland steatosis to steatohepatitis, often termed the 'second hit', remain poorly understood [5-7]. One potential biochemical mechanism of pathogenesis of NASH is oxidative stress [8,9]. Oxidative stress is a cardinal feature of alcoholic steatohepatitis [10]. Histological similarity between NASH and alcohol induced liver disease suggests some shared pathogenetic features such as oxidative stress and cytokine-mediated injury [5-8,11,12]. In addition, evidence of oxidative stress have been found in human livers showing steatosis or NASH [13], and in experimental models of NASH [14]. The efficacy of several antioxidant agents on hepatic steatosis, inflammation, and fibrosis in subjects with NASH has been investigated in several small, open-label studies [15-17]. In all of these studies, those antioxidant agents exerted beneficial effects in improving necroinflammatory activity or fibrosis or both. These findings are also suggestive for the role of oxidative stress in the pathogenesis of NASH.
Various methods have been developed for the measurement of total antioxidant status. However, there is not yet an accepted "gold standard" reference method [18], and decisions concerning standardization, and the terms and units used for the measurement of total antioxidant response (TAR) [19]. This implies that this topic needs to be studied further [20].
In the present study, we aimed to measure TAR using a novel automated method in NASH subjects [20]. As a reciprocal measure, the total peroxide levels were also measured at the same plasma samples. The ratio percentage of the total plasma peroxide level to the plasma TAR value was regarded as oxidative stress index (OSI) [21].
Methods
Enrollment of patients
Twenty-two subjects with biopsy proven NASH (19 male, 3 female; mean age, 37.7 ± 8.8) and 22 healthy controls (17 male, 5 female; mean age, 34.6 ± 9.3) were enrolled in the present study. The study protocol was approved by the local research committee for ethics. All subjects were informed about the study and the written consent was obtained from each one.
Initial evaluation
The major indications for liver biopsy in those 22 subjects were ultrasonographically diagnosed fatty liver and elevation in alanine aminotransferase (ALT).
Diagnosis of NASH was made according to the following criteria:
- Existence of hepatic steatosis (≥10% of hepatocytes affected) with acinar zone 3 hepatocellular injury (ballooning degeneration) and/or lobular inflammation, with or without Mallory's hyaline and pericellular and/or sinusoidal fibrosis, on liver byopsy, which was performed within the previous 12 months. Necroimflamatory grading and fibrosis scoring were based on a modification of the scoring system proposed by Brunt et al [3].
- Negative serological markers for viral infection such as HBsAg, anti-HCV or anti-HIV, and immunological disorders such as antinuclear anti-bodies, anti-smooth muscle antibodies, and anti-liver/kidney microsomes type 1 antibody.
- No evidence that favor of metabolic liver disease such as Wilson's disease and hemochromatosis and alpha-1 antitrypsin deficiency.
Necroimflamatory grades and fibrosis scores of the 22 subjects with NASH were as follows:
- Grade 1 (mild) necroimflamatory changes, in 5 (22.73%) subjects; grade 2 necroimflamatory changes (moderate), in 12 (54.54%) subjects; grade 3 necroimflamatory changes, in 5 (22.73%) subjects.
- Stage 0 Zone 3 perisinusoidal/pericellular fibrosis, in 2 (9.1%) subjects; stage 1, in 9 (40.9%) subjects; stage 2, in 9 (40.9%) subjects; stage 3, in 2 (9.1%) subjects.
- None of the subjects with NASH had stage 4 fibrosis.
Exclusion criteria
Exclusion criteria included recent gastrointestinal by-pass surgery, pregnancy, serum total bilirubin level higher than 2 mg/dL, usage of esterogens, tamoxifen, glucocoticoids, usage of supplemental vitamins, calcium-channel blockers, aspirin, amiodarone, and methotrexate in the previous 6 months, existence of diabetes mellitus, coronary artery disease, rheumatoid arthritis, cancer, systemic or local infection, and history of excess alcohol ingestion, averaging more than 30 gm/day (3 drinks per day) in the previous 10 years, or history of alcohol intake averaging greater than 10 gm/day (1 drink per day: 7 drinks per week) in the previous 1 year.
Recently, it has been suggested that oxidative stress itself may play a crucial role in pathogenesis of diabetes mellitus [22]. Additionally, both increase in oxidative stress and decrease in antioxidant capacities could be related to the complications of diabetes mellitus [23]. Thus, in the present study, we excluded the subjects with diabetes mellitus in order to reflect the accurate role of oxidative stress in pathogenesis NASH.
The clinical and demographic data of the study groups were shown in Table 1.
Blood collection
Blood samples were obtained in fasting state. Samples were withdrawn from a cubital vein into heparinised tubes and immediately stored on ice at 4°C. Then, the plasma was separated from the cells by centrifugation at 3000 rpm for 10 min, and the plasma samples were stored at -80°C until analysis.
Measurement of the total antioxidant status of plasma
The total antioxidant status of the plasma was measured using a novel automated colorimetric measurement method for TAR developed by Erel [20]. In this method, the hydroxyl radical, the most potent biological radical, is produced by the Fenton reaction, and reacts with the colourless substrate O-dianisidine to produce the dianisyl radical, which is bright yellowish-brown in colour. Upon the addition of a plasma sample, the oxidative reactions initiated by the hydroxyl radicals present in the reaction mix are suppressed by the antioxidant components of the plasma, preventing the colour change and thereby providing an effective measure of the total antioxidant capacity of the plasma. The assay results are expressed as mmol Trolox eq./L, and the precision of this assay is excellent, being lower than 3% [24].
Measurement of total plasma peroxide concentration
The total plasma peroxide concentrations were determined using the FOX2 method [25] with minor modifications [21]. The FOX2 test system is based on the oxidation of ferrous iron to ferric iron by the various types of peroxides contained in the plasma samples, in the presence of xylenol orange which produces a coloured ferric-xylenol orange complex whose absorbance can be measured. The FOX2 reagent was prepared by dissolving ammonium ferrous sulphate (9.8 mg) in 250 mM H2SO4 (10 ml) to give a final concentration of 250 mM ferrous iron in acid. This solution was then added to 90 ml HPLC-grade methanol containing 79.2 mg butylated hydroxytoluene (BHT). Finally, 7.6 mg xylenol orange was added, with stirring, to make the working reagent (250 mM ammonium ferrous sulphate, 100 mM xylenol orange, 25 mM H2SO4, and 4 nM BHT, in 90% (v/v) methanol in a final volume of 100 ml). The blank reagent contained all the components of the solution except ferrous sulphate. Aliquots (200 mL) of plasma were mixed with 1.8 ml FOX2 reagent. After incubation at room temparature for 30 min, the vials were centrifuged at 12,000 g for 10 min. The absorbance of the supernatant was then determined at 560 nm. The total peroxide content of the plasma samples was determined as a function of the difference in absorbance between the test and blank samples using a solution of H2O2 as standard. The coefficient of variation for individual plasma samples was less than 5%.
Oxidative stress index
The ratio percentage of the total peroxide to the total anti-oxidant potential gave the oxidative stress index, an indicator of the degree of oxidative stress [21].
Statistical analysis
Data were presented as mean ± SD. Continuous variables were compared using Student t test. Qualitative variables were tested by Chi-square test. Spearman correlation test was used to find out the correlation of fibrosis scores and necroinflamatory grades with total preroxide level, OSI or TAR. Correlations of serum triglyceride and cholesterol levels with total peroxide level, OSI or TAR were analyzed using Pearson correlation test. p < 0,05 was considered as statistical significance.
Results
Two groups were comparable for age, BMI, gender distribution and smoking habit (p value was >0.05 for each comparison). Serum triglyceride and cholesterol levels were significantly higher in subjects with NASH than controls (p value was >0.05 for each comparison).
TAR was 0.85 + 0.11 and 1.88 + 0.32 mmol Trolox eq./L in subjects with NASH and controls, respectively (p < 0.05). Total plasma peroxide level of subjects with NASH and controls was 53.3 + 7.7 and 33.0 + 14.2 μmol H2O2/L, respectively (p < 0.05). OSI was 0.64 + 0.14 and 0.19 + 0.11 AU in subjects with NASH and controls, respectively (p < 0.05).
Total peroxide level and OSI were positively correlated with serum triglyceride level in subjects with NASH (p < 0.05, r = 0.432; p < 0.05, r = 0.521, respectively), while no correlation with serum cholesterol level (p > 0.05 for each comparison). Serum triglyceride and cholesterol levels were not correlated with both total peroxide level and OSI value in control subjects (p > 0.05 for each comparison). No correlation was observed between TAR, and serum triglyceride and cholesterol level in both NASH subjects and controls (p > 0.05 for each comparison).
Fibrosis scores of subjects with NASH were positively correlated with total peroxide level and OSI (p < 0.05, r = 0.607 and p < 0.05, r = 0.728, respectively) (Fig 1, 2), while negatively correlated with TAR (p < 0.05, r = -0.506) (Fig 3). No significant correlation was observed between necroimflamatory grades, and total peroxide level, TAR and OSI (p value was >0.05 for each comparison).
Discussion
Steatosis of the liver is mediated by a multiple factors such as increase in dietary fat intake, release of free fatty acids (FFA) from adipose tissue, insufficient hepatic lipid secretion, and development of insulin resistance [26]. FFA oxidation, CYP2E1 induction, leukocyte infiltration and activation of NADPH oxidase, and mitochodrial dysfunction involving electron transfer inhibition in the respiratory chain increase the production of reactive oxygen species (ROS) such as singlet oxygen, superoxide anion, hydrogen peroxide, and hydroxyl radical [10].
In the present study, in subjects with NASH, significant increases in both total peroxide level and OSI were accompanied with significant decreases in TAR compared to control subjects. In the presence of increased total peroxide level, a decrease in TAR indicates to a high degree of oxidative stress. In addition, in subjects with NASH, total peroxide level and OSI showed positive correlation with fibrosis score. In contrast, TAR showed negative correlation with TAR. On the basis of these findings, it can also be suggested that oxidative stress may have a role in the progression of steatohepatitis to liver fibrosis.
In order to reflect the true state of oxidative stress in the liver, it is more ideal to measure lipid peroxidation markers and antioxidant components in hepatic tissue. Nevertheless, it is impossible to perform multiple tests on very limited amounts of biopsy specimen that obtained in needle biopsy. In addition, liver biopsy carries a significant morbidity and even mortality risk. It is not an ethical approach to perform liver biopsy in healthy controls to provide a comparison with NASH subjects. Thus, in the present study, we have chosen to measure TAR and total peroxide level in plasma samples.
In a few studies, in NASH subjects, antioxidant and oxidant capacities have been investigated [27-29]. Koruk et al [27] reported an increase in oxidative stress in NASH subjects. However, in their study, plasma anitoxidant capacities observed to be comparable in NASH subjects and controls. They suggested that impaired antioxidant defense mechanisms in responding to increased oxidative stress might be an important factor in the pathogenesis of NASH. In a study of Fierbinteanu-Braticevici et al [28], serum index of oxidative stress have been suggested as an independent risk factor for fibrosis in the course of NASH.
In a recent study, which was conducted by Videla et al [29], it has been shown that NAFLD patients with steatosis exhibit a substantial pro-oxidant condition in the liver at early stages of steatosis. This pro-oxidant condition was observed to occur concomitantly with a significant decrease in hepatic SOD activity, changes involving an overall derangement in the antioxidant status of the liver, with the consequent diminution in the antioxidant capacity of plasma. They also observed that further exacerbation in oxidative stress was associated with CYP2E1 induction in patients with steatohepatitis.
Various antioxidants have additive effects on oxidative status. In human serum, the cooperation of these antioxidants protects the organism against the attacks by free radicals [30]. Although the concentration of plasma antioxidant components can be measured individually, these measurements may be time- and cost-consuming and labour intensive. In addition, it may not accurately reflect the total antioxidant status [24]. Thus, the accurate antioxidant capacity of the organism can only be determined by the measurement of TAR [19,20].
In concordance with the study of Videla et al [29], in the present study, we determined the antioxidant capacity using TAR measurement. The oxidants and antioxidant capacity were determined simultaneously to evaluate oxidative stress accurately. However, in study of Videla et al. [29], TAR measurement was performed using ferric reducing ability of plasma (FRAP), while a novel automated method in our study.
The most widely used methods for TAR measurement are colorimetric, fluorescence, and chemiluminescence [31-33]. The fluorescence and chemiluminescence methods need sophisticated techniques and, in most routine clinical biochemistry laboratories, these improved systems are not present. Randox- total antioxidant status (TAS) assay and FRAP assay are the widely used colorimetric TAR measurement methods. In the FRAP assay, the reference range of serum TAR is lowest because this assay practically measures non-protein total antioxidant capacity. However, proteins constitute the main antioxidant component of serum (plasma). The Randox-TAS assay can determine the antioxidative effects of bilirubin, vitamin C, uric acid, polyphenols, and proteins; hence, the reference range for serum TAR is higher than that for the FRAP assay. The novel method is more sensitive for determining the antioxidative effects of bilirubin, uric acid, vitamin C, polyphenols, and proteins. Hence, serum TAR measured by the novel method is higher than those of the Randox-TAS and FRAP assays [20].
In addition, the novel method, which was used in the present study, provides further several advantages in comparison with other currently available methods. It is simple and cheap, and can easily be fully automated. It also does not interact with commonly occurring serum components such as bilirubin, serum lipids, and anticoagulants. Accurate measurements of TAR can be obtained within approximately 10 minutes, making this assay eminently suitable for the clinical biochemistry laboratory [20].
Conclusion
NASH is associated with increased oxidant capacity, especially in the presence of liver fibrosis. The novel automated calorimetric assay is a useful, reliable, simple and easily applicable method in the assessment of the total plasma antioxidant response in NASH.
Competing interests
The author(s) declare that they have no competing intersts.
Authors' contributions
MH conceived the study, and participated in its design and coordination, and in the sequence alignment and drafted the manuscript. CB conceived the study, and participated in its design and coordination and in the sequence alignment, and drafted the manuscript.
FFB conceived the study, and participated in the sequence alignment and drafted the manuscript. TS conceived the study, and participated in its design and coordination. MA conceived the study, and participated in its design and coordination, and collected the samples. SS conceived the study, and participated in its design and coordination, and collected the samples NG conceived the study, collected the samples and carried out the laboratory analysis. OE conceived the study, and participated in its design and coordination, and carried out the laboratory analysis.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Figures and Tables
Figure 1 Correlation between serum total peroxide levels and the score of liver fibrosis among nonalcoholic steatohepatitis group.
Figure 2 Correlation between oxidative stress index and the score of liver fibrosis among nonalcoholic steatohepatitis group. AU, arbitrary unit.
Figure 3 Correlation between serum total antioxidant response and the score of liver fibrosis among nonalcoholic steatohepatitis group.
Table 1 The clinical and demographic data of the study groups
Subjects with NASH Control subjects
Age 37.7 ± 8.8 34.6 ± 9.3
Weight 83.5 ± 9.4 80.7 ± 14.5
Height 171.1 ± 5.8 169 ± 7.8
BMI 28.9 ± 3 27.7 ± 3.3
Triglyceride 228 ± 93.7* 159.8 ± 62.6
Cholesterol 205.4 ± 44.4* 168.5 ± 35.1
ALT 70 ± 29 23 ± 6
Smoking habit 14 (63.6%) 11 (50%)
Data were presented as mean ± SD.
*p<0.05 vs. controls.
ALT, Alanine Aminotransferase; BMI, Body Mass Index.
==== Refs
Angulo P Nonalcoholic fatty liver disease N Engl J Med 2002 346 1221 31 11961152 10.1056/NEJMra011775
Neuschwander-Tetri BA Caldwell SH Nonalcoholic steatohepatitis: summary of an AASLD Single Topic Conference Hepatology 2003 37 1202 1219 12717402 10.1053/jhep.2003.50193
Brunt EM Janney CG Di Bisceglie AM Neuschwander-Tetri BA Bacon BR Nonalcoholic steatohepatitis: a proposal for grading and staging the histological lesions Am J Gastroenterol 1999 94 2467 2474 10484010 10.1111/j.1572-0241.1999.01377.x
Matteoni CA Younossi ZM Gramlich T Boparai N Liu YC McCullough AJ Nonalcoholic fatty liver disease: a spectrum of clinical and pathological severity Gastroenterology 1999 116 1413 1419 10348825 10.1016/S0016-5085(99)70506-8
Caldwell SH Oelsner DH Iezzoni JC Hespenheide EE Battle EH Driscoll CJ Cryptogenic cirrhosis: clinical characterization and risk factors for underlying disease Hepatology 1999 29 664 669 10051466 10.1002/hep.510290347
Green RM NASH – hepatic metabolism and not simply the metabolic syndrome Hepatology 2003 38 14 17 12829980 10.1053/jhep.2003.50325
Yang SQ Lin HZ Lane MD Clemens M Diehl AM Obesity increases sensitivity to endotoxin liver injury: implications for the pathogenesis of steatohepatitis Proc Natl Acad Sci U S A 1997 94 2557 2562 9122234 10.1073/pnas.94.6.2557
Chitturi S Farrell GC Etiopathogenesis of nonalcoholic steatohepatitis Semin Liver Dis 2001 21 27 41 11296694 10.1055/s-2001-12927
Day CP James OF Steatohepatitis: a tale of two "hits"? Gastroenterology 1998 114 842 845 9547102 10.1016/S0016-5085(98)70599-2
Videla LA Rodrigo RN Araya J Poniachik J Oxidative stress and depletion of hepatic long-chain polyunsaturated fatty acids may contribute to nonalcoholic fatty liver disease Free Radic Biol Med 2004 37 1499 1507 15454290 10.1016/j.freeradbiomed.2004.06.033
Su GL Lipopolysaccharides in liver injury: molecular mechanisms of Kupffer cell activation Am J Physiol Gastrointest Liver Physiol 2002 283 G256 265 12121871
Hoek JB Pastorino JG Ethanol, oxidative stress, and cytokine-induced liver cell injury Alcohol 2002 27 63 68 12062639 10.1016/S0741-8329(02)00215-X
Bhat VB Madyastha KM Antioxidant and radical scavenging properties of 8-oxo derivatives of xanthine drugs pentoxifylline and lisofylline Biochem Biophys Res Commun 2001 288 1212 1217 11700041 10.1006/bbrc.2001.5922
Leclercq IA Farrell GC Field J Bell DR Gonzalez FJ Robertson GR CYP2E1 and CYP4A as microsomal catalysts of lipid peroxides in murine nonalcoholic steatohepatitis J Clin Invest 2000 105 1067 1075 10772651
Hasegawa T Yoneda M Nakamura K Makino I Terano A Plasma transforming growth factor-beta1 level and efficacy of alpha-tocopherol in patients with non-alcoholic steatohepatitis: a pilot study Aliment Pharmacol Ther 2001 15 1667 1672 11564008 10.1046/j.1365-2036.2001.01083.x
Sanyal AJ Mofrad PS Contos MJ Sargeant C Luketic VA Sterling RK Stravitz RT Shiffman ML Clore J Mills AS A pilot study of vitamin E versus vitamin E and pioglitazone for the treatment of nonalcoholic steatohepatitis Clin Gastroenterol Hepatol 2004 2 1107 1115 15625656 10.1016/S1542-3565(04)00457-4
Harrison SA Torgerson S Hayashi P Ward J Schenker S Vitamin E and vitamin C treatment improves fibrosis in patients with nonalcoholic steatohepatitis Am J Gastroenterol 2003 98 2485 2490 14638353 10.1111/j.1572-0241.2003.08699.x
Hubel CA Oxidative stress in the pathogenesis of preeclampsia Proc Soc Exp Biol Med 1999 222 222 235 10601881 10.1046/j.1525-1373.1999.d01-139.x
Erel O A novel automated direct measurement method for total antioxidant capacity using a new generation, more stable ABTS radical cation Clin Biochem 2004 37 277 285 15003729 10.1016/j.clinbiochem.2003.11.015
Erel O A novel automated method to measure total antioxidant response against potent free radical reactions Clin Biochem 2004 37 112 119 14725941 10.1016/j.clinbiochem.2003.10.014
Harma M Harma M Erel O Increased oxidative stress in patients with hydatidiform mole Swiss Med Wkly 2003 133 563 566 14691728
Davi G Falco A Patrono C Lipid peroxidation in diabetes mellitus Antioxid Redox Signal 2005 7 256 268 15650413 10.1089/ars.2005.7.256
Gil-Del Valle L Milian Lde L Toledo A Vilaro N Tapanes R Otero MA Altered redox status in patients with Diabetes Mellitus type I Pharmacol Res 2005 51 375 380 15683752 10.1016/j.phrs.2004.10.012
Cao G Prior RL Comparison of different analytical methods for assessing total antioxidant capacity of human serum Clin Chem 1998 44 1309 1315 9625058
Miyazawa T Determination of phospholipid hydroperoxides in human blood plasma by a chemiluminescence-HPLC assay Free Radic Biol Med 1989 7 209 217 2806945 10.1016/0891-5849(89)90017-8
Te Sligte K Bourass I Sels JP Driessen A Stockbrugger RW Koek GH Non-alcoholic steatohepatitis: review of a growing medical problem Eur J Intern Med 2004 15 10 21 15066643 10.1016/j.ejim.2003.12.008
Koruk M Taysi S Savas MC Yilmaz O Akcay F Karakok M Oxidative stress and enzymatic antioxidant status in patients with nonalcoholic steatohepatitis Ann Clin Lab Sci 2004 34 57 62 15038668
Fierbinteanu-Braticevici C Bengus A Neamtu M Usvat R The risk factors of fibrosis in nonalcoholic steatohepatitis Rom J Intern Med 2002 40 81 88 15526543
Videla LA Rodrigo R Orellana M Fernandez V Tapia G Quinones L Varela N Contreras J Lazarte R Csendes A Rojas J Maluenda F Burdilas P Diaz JC Smok G Thielemann L Poniachik J Oxidative stress-related parameters in the liver of non-alcoholic fatty liver disease patients Clin Sci (Lond) 2004 106 261 268 14556645 10.1042/CS20030285
Wayner DD Burton GW Ingold KU Barclay LR Locke SJ The relative contributions of vitamin E, urate, ascorbate and proteins to the total peroxyl radical-trapping antioxidant activity of human blood plasma Biochim Biophys Acta 1987 924 408 419 3593759
Janaszewska A Bartosz G Assay of total antioxidant capacity: comparison of four methods as applied to human blood plasma Scand J Clin Lab Invest 2002 62 231 236 12088342 10.1080/003655102317475498
Prior RL Cao G In vivo total antioxidant capacity: comparison of different analytical methods Free Radic Biol Med 1999 27 1173 1181 10641708 10.1016/S0891-5849(99)00203-8
Schlesier K Harwat M Bohm V Bitsch R Assessment of antioxidant activity by using different in vitro methods Free Radic Res 2002 36 177 187 11999386 10.1080/10715760290006411
|
16283935
|
PMC1308812
|
CC BY
|
2021-01-04 16:03:28
|
no
|
BMC Gastroenterol. 2005 Nov 11; 5:35
|
utf-8
|
BMC Gastroenterol
| 2,005 |
10.1186/1471-230X-5-35
|
oa_comm
|
==== Front
BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-1671630575310.1186/1471-2164-6-167Research ArticleAnalysis of vertebrate genomes suggests a new model for clade B serpin evolution Kaiserman Dion [email protected] Phillip I [email protected] Department of Biochemistry & Molecular Biology, Monash University, Clayton, Victoria, Australia2005 23 11 2005 6 167 167 16 9 2005 23 11 2005 Copyright © 2005 Kaiserman and Bird; licensee BioMed Central Ltd.2005Kaiserman and Bird; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The human genome contains 13 clade B serpin genes at two loci, 6p25 and 18q21. The three genes at 6p25 all conform to a 7-exon gene structure with conserved intron positioning and phasing, however, at 18q21 there are two 7-exon genes and eight genes with an additional exon yielding an 8-exon structure. Currently, it is not known how these two loci evolved, nor which gene structure arose first – did the 8-exon genes gain an exon, or did the 7-exon genes lose one? Here we use the genomes of diverse vertebrate species to plot the emergence of clade B serpin genes and to identify the point at which the two genomic structures arose.
Results
Analysis of the chicken genome indicated the presence of a single clade B serpin gene locus, containing orthologues of both human loci and both genomic structures. The frog genome and the genomes of three fish species presented progressively simpler loci, although only the 7-exon structure could be identified. The Serpinb12 gene contains seven exons in the frog genome, but eight exons in chickens and humans, indicating that the additional exon evolved in this gene.
Conclusion
We propose a new model for clade B serpin evolution from a single 7-exon gene (either Serpinb1 or Serpinb6). An additional exon was gained in the Serpinb12 gene between the tetrapoda and amniota radiations to produce the 8-exon structure. Both structures were then duplicated at a single locus until a chromosomal breakage occurred at some point along the mammalian lineage resulting in the two modern loci.
==== Body
Background
The serpins are a superfamily of proteins sharing a conserved tertiary structure [1] that has evolved primarily to control proteolytic activity, although some have evolved non-inhibitory functions [2-6]. To date, over 500 serpin sequences have been identified in the genomes of species from almost all phyla including viruses, bacteria, metazoans and plants [1,7].
The serpin mechanism of inhibition is based on the metastable nature of the native serpin fold [8]. The reactive centre loop (RCL) is exposed at the top of the serpin where it acts as a pseudosubstrate for the target proteinase [9]. Following cleavage of the RCL, a conformational rearrangement, termed the stressed to relaxed transition, occurs and the RCL inserts into a β-sheet forming an extra β-strand [10]. This rearrangement is responsible for the distortion and irreversible inactivation of the target proteinase [11].
Phylogenetic analysis has divided serpins into 16 clades and a number of orphan sequences [1]. The clade B serpins are primarily intracellular proteinase inhibitors [12]. In humans this clade consists of 13 members and includes regulators of inflammation, apoptosis and angiogenesis (reviewed in [13]). The genes are arranged at two genomic loci at 6p25 (3 genes: SERPINB1, SERPINB6 and SEPINB9) and 18q21 (10 genes: SERPINB2, B3, B4, B5, B7, B8, B10, B11, B12 and B13), and all clade B serpins described to date conform to two related gene structures of seven or eight exons. Both the phasing and positioning of the intron/exon splice sites are identical for six introns, leading to conserved exon lengths. The additional intron splice site of the 8-exon structure is also conserved, although the extra exon encodes a highly divergent sequence of variable length, termed the CD loop.
Although clade B serpins have been identified in mammals (humans, mice and rats) as well as birds, it is not known how they evolved, nor has the primordial structure or gene been identified. Two competing models exist to explain the presence of two clade B loci in humans. The first proposes duplication of the entire 6p25 locus followed by a number of single gene duplications at 18q21 [14]. The second proposes duplication of a single gene from 18q21 to 6p25, followed by successive duplications at both loci to derive the modern complement [15]. Two key differences are apparent between these models. The former assumes that the 7-exon structure and 6p25 locus are ancestral, whereas the latter assumes the opposite.
Here, we use the completed chicken genome to show that the two mammalian loci are the result of an ancient chromosomal split, as orthologues of genes from both human loci are linked in the chicken. Further analysis of emerging genome sequences suggests that clade B serpin genes arose from a primordial Serpinb1 or b6 gene with seven exons, the additional exon being gained in Serpinb12 after the divergence of amphibians and amniotes.
Results
Clade B serpin genes in the chicken (Gallus gallus)
To determine which of the two clade B serpin loci is ancestral, the chicken genome was investigated, as it represents a distantly related species. This analysis was also recently performed independently by another group, yielding similar results [16]. As illustrated in Figure 2A, the chicken genome contains a single clade B serpin locus on chromosome 2, composed of 10 genes, nine of which are supported by EST data. Both gene structures are also evident within the locus. Single orthologues exist in the chicken for the 7-exon human genes SERPINB1 (MNEI), SERPINB5 (maspin) and SERPINB6 (PI6), and also for the 8-exon genes SERPINB2 (PAI-2) and SERPINB12 (yukopin). Human SERPINB10 (bomapin), an 8-exon gene, has 2 chicken homologues – MENT (which fits the 8-exon arrangement) and a MENT-like gene for which the structure cannot be completely determined. The three remaining chicken genes Serpinb14 (ovalbumin), Serpinb14b (geneY) and Serpinb14c (geneX) are highly related [17] and have no human orthologues. This suggests that the common ancestor of chickens and mammals had a complement of six clade B serpin genes: SERPINB1, B2, B5, B6, B10 and B12. Furthermore, the organization of the chicken clade B serpin locus matches that of humans, contradicting both previously proposed models of clade B serpin gene evolution. Rather than two independently evolving loci, it suggests that both the 7- and 8-exon gene structures evolved at a single locus in the common ancestor of birds and mammals that then split to yield the 2 loci apparent in humans and rodents.
Figure 1 Distribution of clade B serpin genes in vertebrate genomes. A dendrogram showing relationships between vertebrate species with clade B serpin gene loci. The genes identified within each genome are indicated on the right without the SERPIN root designation. Multiple loci are delineated by square brackets, while genes identified in EST databases, but not in the genome, are in round brackets. * due to the incomplete nature of the sequencing data, the structure of the rabbit (Oryctolagus cuniculus) loci could not be established. Branch lengths are not to scale.
Figure 2 Comparison of frog, chicken and human clade B serpin loci. The structure of the serpin locus in frog, chicken and human is shown with 7-exon serpin genes in green, 8-exon serpin genes in red and non-serpin genes in black. Gene structure could not be predicted for Serpinb10a (blue). Arrows indicate the direction of transcription. Orthologous genes with strong RCL conservation are linked with solid lines, dotted lines denote inter-species orthologues with weak RCL homology. (B) The amino acid sequences of human, chicken and frog clade B serpins were aligned with human antithrombin (SERPINC1) and a Neighbour-Joining tree constructed (gapped positions removed, 1000 bootstraps) to show evolutionary relationships. Scale bar indicates the number of substitutions at each site. Node colour indicates the bootstrap value (black >75%, grey 50–75%, white <50%). Genes are named without the SERPIN root. hsa = Homo sapiens; gga = Gallus gallus; xtr =
Xenopus tropicalis.
Evolution of clade B serpin genes in vertebrates
Since both the 7 and 8-exon gene structures were present in the chicken genome, we investigated more distantly related genomes to identify which structure arose first. Each of the 13 human protein sequences were used to probe the current builds of vertebrate genomes to identify clade B serpin genes. This identified genes with strong RCL similarity to human clade B serpins in the genomes of fish, amphibians and various mammals. By plotting the presence of genes on a dendrogram of species a pattern of gene evolution can be proposed (Figure 1).
As described above, the common ancestor of mammals and birds contained a single clade B serpin gene locus of six genes (Serpinb1, b2 b5, b6, b10 and b12). All of the mammalian genomes examined so far (human, chimpanzee, dog, rat and mouse) contain two loci, suggesting that the chromosomal split occurred early in mammalian evolution (Figure 1). The locus syntenic with human 6p25 contains SERPINB1, B6 and B9 in humans, chimpanzees, and dogs and an expanded repertoire of 15 and 8 in mice and rats, respectively [18,19], indicating further expansion within the rodent lineage. The incomplete rabbit genome yielded hits for 10 of the human clade B serpins (Serpinb3, b4 and b10 were not identified), with no evidence of the expansion observed in rats or mice. Furthermore, the Serpinb9 RCL sequence of rabbits is predicted to match that of humans whereas the rodent Serpinb9 RCL is mutated.
The locus syntenic to human 18q21 is identical between mammals with the exception of the SERPINB3 and SERPINB4 genes (SCCA-1 and -2), which are unique to each species (data not shown and [20]). This suggests that 12 genes were present in the mammalian common ancestor – SERPINB1, B6 and B9 at one locus, and SERPINB2, B5, B7, B8, B10, B11, B12, B13 and the primordial SERPINB3/B4 at the other. Since then, the ancestral SERPINB3/B4 has evolved independently in each mammalian species.
Further back in evolution, chickens and mammals shared a common ancestor with amphibians approximately 300 million years ago, and the genome of Xenopus tropicalis (the western clawed frog) provides an insight into this more primitive gene cluster (Figure 1). The X. tropicalis genome contains a single clade B serpin locus of 4 genes (Serpinb1, Serpinb6, Serpinb5 and Serpinb12), bounded by orthologues of WHIP and VPS4B. The placement of both serpin and non-serpin genes is conserved between the frog and chicken genomes, strengthening our hypothesis that a single clade B serpin gene cluster has split at some point in mammalian evolution (Figure 2A). Again, the phylogenetic tree confirms that clade B serpin genes cluster on the basis of conserved function, rather than by species (Figure 2B).
450 million years ago, amphibians, birds and mammals shared a common ancestor with fish [21]. Three fish genomes were available for analysis: Danio rerio (zebrafish), Tetraodon nigroviridis (green pufferfish) and Takifugu rubripes (japanese pufferfish). The T. rubripes and T. nigroviridis genomes each contain one or two copies of Serpinb1, respectively, with no other serpin genes for over 100 kb upstream or downstream, while the D. rerio locus contains two copies of Serpinb1, linked to two other genes with RCL sequences that do not match any human serpins. However, ESTs derived from both Serpinb1 and Serpinb6 can be identified from all three species, indicating the presence of another clade B serpin locus. Indeed, a second locus is identifiable in the D. rerio and T. nigroviridis genomes. The T. nigroviridis locus contains only Serpinb6, although the contig is very short and other genes may exist. The D. rerio region contains a number of clade B serpin gene fragments as well as a gene with RCL homology to human SERPINB12, although no orthologue of SERPINB6 can be identified. However, the BAC clone finishes in intronic sequence between the final two exons of a serpin gene, and therefore this fragment may be the 5' region of D. rerio Serpinb6.
Gene structure
The gene structures of the newly identified genes from Xenopus and fish were determined by alignment of ESTs to genomic sequence as well as by GENSCAN predictions (Table 2). Human SERPINB1 is a 7-exon gene, and this is also the case for Serpinb1 in the three fish genomes investigated, as well as the novel genes identified in Danio rerio. Lack of EST coverage resulted in an inability to determine intron phasing for the T. nigroviridis and T. rubripes genes, however, the positioning of the introns within the predicted mRNA was consistent with the mammalian clade B serpins.
Table 1 Accession numbers of sequences used in this study
Gene (Species) Accession Number(s)
Serpinb1 (Oncorhynchus mykiss) [Genbank:CX143380]
Serpinb6 (Oncorhynchus mykiss) [Genbank:AY606039]
Serpinb1 (Salmo salar) [Genbank:CA059107; Genbank:CA051086]
Serpinb6 (Salmo salar) [Genbank:CK877893]
Serpinb1 (Oryzias latipes) [Genbank:BJ721045]
Serpinb6 (Oryzias latipes) [Genbank:BJ530676]
Serpinb6 (Ictalurus punctatus) [Genbank:CK420790]
Serpinb1 (Haplochromis) [Genbank:BJ692672]
Serpinb1 (Gasterosteus aculeatus) [Genbank:CD504030]
Serpinb1 (Danio rerio) [Genbank:AL912530]
Serpinb6 (Danio rerio) [Genbank:CD015363]
Serpinb1 (Xenopus tropicalis) [Genbank:CX372321; Genbank:CF593092; Genbank:CX383158; Genbank:DR857859]
Serpinb5 (Xenopus tropicalis) [Genbank:CX847450; Genbank:CF346211]
Serpinb6 (Xenopus tropicalis) [Genbank:BX709151; Genbank:DR860385; Genbank:DN093291]
Serpinb12 (Xenopus tropicalis) [Genbank:DR880497]
Serpinb1 (Gallus gallus) [Genbank:CD216625; Genbank:CF255528; Genbank:CF257148]
Serpinb2 (Gallus gallus) [Genbank:BU409199; Genbank:BU205882; Genbank:BU410099]
Serpinb5 (Gallus gallus) [Genbank:CD739917]
Serpinb6 (Gallus gallus) [Genbank:CF250725; Genbank:BU116697; BU133732; Genbank:AJ734077]
Serpinb12 (Gallus gallus) [Genbank:BU442268]
Table 2 Intron/exon phasing in vertebrate clade B serpin genes
Species Gene1 A2 B C3 D E F G
D. rerio B1 UTR 0 NP 0 1 0 0
T. nigroviridis†
T. rubripes† B1 UTR 0 NP 0 1 0 0
X. tropicalis B1 UTR 0 NP 0 1 0 0
B5 UTR 0 NP 0 1 0 0
B6 UTR 0 NP 0 1 0 0
B12 UTR 0 NP 0 1 0 0
G. gallus B1 UTR 0 NP 0 1 0 0
B2 UTR 0 0 0 1 0 0
B5 UTR 0 NP 0 1 0 0
B6 UTR 0 NP 0 1 0 0
B10a UTR 0 * 0 1 0 0
B10b UTR 0 0 0 1 0 0
B12 UTR 0 0 0 1 0 0
B14 UTR 0 0 0 1 0 0
B14b UTR 0 0 0 1 0 0
B14c UTR 0 0 0 1 0 0
H. Sapiens B1 UTR 0 NP 0 1 0 0
B2 UTR 0 0 0 1 0 0
B3/4 UTR 0 0 0 1 0 0
B5 UTR 0 NP 0 1 0 0
B6 UTR 0 NP 0 1 0 0
B7 UTR 0 0 0 1 0 0
B8 UTR 0 NP 0 1 0 0
B9 UTR 0 NP 0 1 0 0
B10 UTR 0 0 0 1 0 0
B11 UTR 0 0 0 1 0 0
B12 UTR 0 0 0 1 0 0
B13 UTR 0 0 0 1 0 0
1 Gene names are given without the Serpin root.
2 UTR = intron occurs in the 5' untranslated region and thus has no phase
3 NP = Intron not present, * = Intron cannot be reliably predicted and no EST exists.
† Intron phasing predicted from consensus sequence
Overlapping ESTs allowed construction of complete mRNA sequences for all 4 of the Xenopus genes. Serpinb1, b5 and b6 conformed to the 7-exon structure, and both positioning and phasing of the intron/exon boundaries was conserved with humans. The other gene to appear in frogs is Serpinb12, an 8-exon gene in the chicken and mammalian genomes. However, this gene in frogs contained only 7-exons, lacking the extra exon encoding the CD loop. This gene structure was confirmed by EST DR880497 which spans all 6 intron/exon boundaries. As shown in Figure 2B, Xenopus Serpinb12 is clearly related to human and chicken Serpinb12, suggesting that the 8-exon structure arose some time between the divergence of amphibians and amniotes in the Serpinb12 gene.
Discussion
The human genome contains two clusters of clade B serpin genes at 6p25 and 18q21. The 13 genes all conform to either a 7- or 8-exon gene structure with conservation of both phasing and positioning of intron/exon boundaries. To address possible mechanisms of clade B gene evolution, the chicken genome was analysed and found to contain a single clade B gene cluster of 10 genes. Another recent investigation of the chicken genome also identified these 10 genes, however, the authors concluded that the newly identified MENT-like chicken gene, Serpinb10a, and human SERPINB10 represented direct orthologues [16]. This was based on the overall protein similarity as well as homology within the CD loop and RCL. While the RCL sequence of chicken Serpinb10a and human SERPINB10 have some similarities, there are a number of charge differences throughout the RCL, and optimal alignment requires insertion of a gap at the P side (N-terminal to the point of proteolytic cleavage), a feature not observed between other chicken-human orthologues. Furthermore, although the presence of a CD loop would be predicted in Serpinb10a given its evolutionary ancestry, the presence of an eighth exon requires experimental proof. Without direct sequencing of a transcript, it remains possible that the exon may be missing or contain a premature stop codon or mis-sense mutation resulting in a lack of functional protein expression. The exon identified by Benarafa and Remold-O'Donnell had a very low statistical score [16], and decreasing the GENSCAN settings to detect this exon leads to the identification of a number of potential exons in other serpin genes that are not transcribed (data not shown).
The identification of orthologues (matching RCLs) of both mammalian clusters (SERPINB2, SERPINB5 and SERPINB12 from human 18q21 and SERPINB1 and SERPINB6 from human 6p25) at a single genomic locus in the chicken raises the intriguing possibility that the serpin clusters arose from an ancient chromosomal breakage, rather than chromosomal duplication events. More distantly related genomes appear to support this conclusion as a single clade B serpin gene cluster exists in the Xenopus tropicalis genome, consisting of Serpinb1, b5, b6 and b12. Therefore, the two loci observed in mammals arose from this single cluster of four 7-exon genes. Serpinb12 gained an extra exon and was duplicated to give rise to the 8-exon genes, followed by a chromosomal split yielding two loci. This presents a much simpler mechanism for the observation of different genetic structures at a single locus than previously proposed models [14,15] (Figure 3).
Figure 3 A new mechanism of clade B serpin gene expansion. (A) A new model for evolution of two clade B serpin loci in mammals, from a single locus of 4 genes. A series of intrachromosomal duplications results in 13 genes, followed by chromosomal breakage to yield two loci. Two previously proposed mechanisms based on duplication of an ancestral locus as described by (B) Bartuski et al. [14]; and (C) Scott et al. [15]. Arrows represent gene duplication events. Black boxes indicate genes displaying the 8-exon structure, white boxes are 7-exon genes. Genes are named without the SERPIN root. Diagram not to scale.
Since Serpinb1 and Serpinb6 are linked in frogs, chickens and humans, it is likely that they were linked in their common ancestor with fish, however, this was not the case in any of the fish genomes investigated. This may be due to a whole genome duplication that occurred early in the evolution of ray finned fish, followed by rapid loss of duplicate genes [22]. This could give rise to the two serpin loci that have since evolved separately (Figure 4). Therefore, we propose that the single primordial serpin clade B locus (in the common ancestor of fish and tetrapods) contained only Serpinb1 and Serpinb6, as these are the only two genes identifiable in ESTs from multiple species of fish. Following the whole genome duplication, one gene was lost from each locus (resulting in one locus with Serpinb1 and another with Serpinb6) and each of these loci have since evolved uniquely in descendant fish species (Figure 4). The fish family Polypteridae (bichirs) diverged from teleost fish before the whole genome duplication [22], and as such, a genomic analysis within this family would be predicted to show a single 'primordial' clade B serpin gene locus of Serpinb1 and Serpinb6.
Figure 4 Distribution of clade B serpin ESTs in fish. A dendrogram of fish species is shown with clade B serpins ESTs shown on the right in brackets. The tetrapoda lineage is also indicated, with the simplest locus (Xenopus tropicalis) shown for comparison. Genes are named without the SERPIN root. The relative timing of genome duplication and subsequent gene loss within ray finned fish are shown as well as the composition of the loci after each event. Asterisk indicates an EST for which only the genus name was given.
The identification of a Serpinb12 homologue in the D. rerio genome suggests that this gene may have evolved significantly earlier than the tetrapod lineage, however, no corroborating ESTs can be found in other fish species. A number of unique clade B serpins have evolved in zebrafish, at both genomic loci, which cannot be found in other fish genomes or EST databases, indicating a more rapid evolution in this fish species. Furthermore, phylogenetic analysis indicates that Danio Serpinb12 does not cluster with Serpinb12 from other species, but rather with Danio Serpinb6 (data not shown), and suggests that the Danio rerio Serpinb12 gene may be a case of convergent evolution.
Six genes (Serpinb3/b4, b7, b8, b9, b11 and b13) were gained early along the mammalian branch and the cluster split into two loci, although when these events occurred in relation to each other cannot currently be discerned. The current effort to sequence the platypus (a monotreme) genome, as well as a number of marsupials (the wallaby and opossum), which branched off early in mammalian evolution may shed some light on this.
The analysis of frog Serpinb12 clearly shows that it is a 7-exon gene, despite having eight exons in both chickens and mammals. This shows that the 7-exon structure arose first and that the eighth exon was gained in Serpinb12 some time after amphibians evolved, but before mammals and birds split. The gain of the CD loop has enabled some members of the clade B serpin family to evolve motifs for secondary functions, without affecting their inhibitory role [23]. The AT-hook motif and nuclear localisation signal (NLS) of chicken MENT [3] are found in the CD-loop, as is the NLS of human bomapin [24]. The extended CD-loop of PAI-2 contains motifs necessary for its suppression of tumour necrosis factor α induced death [23], redox sensitivity [25,26] and cross-linking to cell membranes [27,28] as well as a novel motif for binding to retinoblastoma protein [29].
These analyses suggest that either Serpinb1 or Serpinb6 is the original clade B serpin gene, and that it was present in the common ancestor of euteleostomi (the bony vertebrates, including both fish and mammals) approximately 450 million years ago. A possible functional orthologue of Serpinb6 is evident in the urochordate species Ciona intestinalis (sea squirt). However, this gene does not contain any introns, although other serpin genes in the Ciona genome do have introns. This makes it difficult to suggest the Ciona Serpinb6 gene as a primordial clade B serpin. Furthermore, the suggestion that clade B serpins evolved from an intronless precursor would also be contrary to a recent study suggesting that they arose from an intron-containing gene [30].
The question of where clade B genes arose may be answered when the genomes of the skate and lamprey are completely sequenced. These species diverged at different times between the sea squirt-human and fish-human common ancestors, and may indicate which gene, Serpinb1 or Serpinb6, evolved first, and whether or not clade B serpins are descendant from an intronless precursor.
Conclusion
By investigating the genomes of various vertebrate species, we have shown that the all clade B serpins are descended from a single gene (either Serpinb1 or Serpinb6) that contained seven exons. The additional exon present in the 8-exon structure appeared in the Serpinb12 gene at some point after the divergence of amphibians from amniotes, but before the divergence of birds and mammals. Therefore, both the 7-exon and 8-exon genes evolved at a single locus that split soon after birds and mammals diverged two yield the two loci present in the human genome.
Methods
The current builds of vertebrate genomes [22,31-34] were accessed through the NCBI and mined using BLAST [35]. The Xenopus tropicalis genome was accessed from the Department of Energy Joint Genome Institute. Rabbit genomic contigs were downloaded from the Broad Institute. GENSCAN [36] was also used to predict genes from genomic sequence. Intron/exon boundaries were predicted with Spidey. Alignments were created using ClustalW [37] and manually adjusted with BioEdit (version 6.0.7). Phylogenetic relationships were determined using the Neighbour-Joining method with 1,000 bootstraps and trees were viewed with TreeView [38]. Accession numbers for the sequences identified in this study are given in Table 1.
Authors' contributions
DK performed the analysis and drafted the manuscript. PIB assisted in design of the study and in drafting of the manuscript.
==== Refs
Irving JA Pike RN Lesk AM Whisstock JC Phylogeny of the serpin superfamily: implications of patterns of amino acid conservation for structure and function Genome Res 2000 10 1845 1864 11116082 10.1101/gr.GR-1478R
Doolittle RF Angiotensinogen is related to the antitrypsin-antithrombin-ovalbumin family Science 1983 222 417 419 6604942
Grigoryev SA Bednar J Woodcock CL MENT, a heterochromatin protein that mediates higher order chromatin folding, is a new serpin family member J Biol Chem 1999 274 5626 5636 10026180 10.1074/jbc.274.9.5626
Zou Z Anisowicz A Hendrix MJ Thor A Neveu M Sheng S Rafidi K Seftor E Sager R Maspin, a serpin with tumor-suppressing activity in human mammary epithelial cells Science 1994 263 526 529 8290962
Hirayoshi K Kudo H Takechi H Nakai A Iwamatsu A Yamada KM Nagata K HSP47: a tissue-specific, transformation-sensitive, collagen-binding heat shock protein of chicken embryo fibroblasts Mol Cell Biol 1991 11 4036 4044 2072906
Hammond GL Smith CL Goping IS Underhill DA Harley MJ Reventos J Musto NA Gunsalus GL Bardin CW Primary structure of human corticosteroid binding globulin, deduced from hepatic and pulmonary cDNAs, exhibits homology with serine protease inhibitors Proc Natl Acad Sci U S A 1987 84 5153 5157 3299377
Irving JA Steenbakkers PJ Lesk AM Op den Camp HJ Pike RN Whisstock JC Serpins in prokaryotes Mol Biol Evol 2002 19 1881 1890 12411597
Carrell RW Owen MC Plakalbumin, alpha 1-antitrypsin, antithrombin and the mechanism of inflammatory thrombosis Nature 1985 317 730 732 3877243 10.1038/317730a0
Stein PE Leslie AG Finch JT Turnell WG McLaughlin PJ Carrell RW Crystal structure of ovalbumin as a model for the reactive centre of serpins Nature 1990 347 99 102 2395463 10.1038/347099a0
Stratikos E Gettins PG Formation of the covalent serpin-proteinase complex involves translocation of the proteinase by more than 70 A and full insertion of the reactive center loop into beta-sheet A Proc Natl Acad Sci U S A 1999 96 4808 4813 10220375 10.1073/pnas.96.9.4808
Huntington JA Read RJ Carrell RW Structure of a serpin-protease complex shows inhibition by deformation Nature 2000 407 923 926 11057674 10.1038/35038119
Remold-O'Donnell E The ovalbumin family of serpin proteins FEBS Lett 1993 315 105 108 8417965 10.1016/0014-5793(93)81143-N
Silverman GA Whisstock JC Askew DJ Pak SC Luke CJ Cataltepe S Irving JA Bird PI Human clade B serpins (ov-serpins) belong to a cohort of evolutionarily dispersed intracellular proteinase inhibitor clades that protect cells from promiscuous proteolysis Cell Mol Life Sci 2004 61 301 325 14770295 10.1007/s00018-003-3240-3
Bartuski AJ Kamachi Y Schick C Overhauser J Silverman GA Cytoplasmic antiproteinase 2 (PI8) and bomapin (PI10) map to the serpin cluster at 18q21.3 Genomics 1997 43 321 328 9268635 10.1006/geno.1997.4827
Scott FL Eyre HJ Lioumi M Ragoussis J Irving JA Sutherland GA Bird PI Human ovalbumin serpin evolution: phylogenic analysis, gene organization, and identification of new PI8-related genes suggest that two interchromosomal and several intrachromosomal duplications generated the gene clusters at 18q21-q23 and 6p25 Genomics 1999 62 490 499 10644448 10.1006/geno.1999.6021
Benarafa C Remold-O'donnell E The ovalbumin serpins revisited: Perspective from the chicken genome of clade B serpin evolution in vertebrates Proc Natl Acad Sci U S A 2005 102 11367 11372 16055559 10.1073/pnas.0502934102
Heilig R Perrin F Gannon F Mandel JL Chambon P The ovalbumin gene family: structure of the X gene and evolution of duplicated split genes Cell 1980 20 625 637 7418002 10.1016/0092-8674(80)90309-8
Kaiserman D Knaggs S Scarff KL Gillard A Mirza G Cadman M McKeone R Denny P Cooley J Benarafa C Remold-O'Donnell E Ragoussis J Bird PI Comparison of human chromosome 6p25 with mouse chromosome 13 reveals a greatly expanded ov-serpin gene repertoire in the mouse Genomics 2002 79 349 362 11863365 10.1006/geno.2002.6716
Puente XS Lopez-Otin C A genomic analysis of rat proteases and protease inhibitors Genome Res 2004 14 609 622 15060002 10.1101/gr.1946304
Askew DJ Askew YS Kato Y Turner RF Dewar K Lehoczky J Silverman GA Comparative genomic analysis of the clade B serpin cluster at human chromosome 18q21: amplification within the mouse squamous cell carcinoma antigen gene locus Genomics 2004 84 176 184 15203215 10.1016/j.ygeno.2004.01.015
Hedges SB The origin and evolution of model organisms Nat Rev Genet 2002 3 838 849 12415314 10.1038/nrg929
Jaillon O Aury JM Brunet F Petit JL Stange-Thomann N Mauceli E Bouneau L Fischer C Ozouf-Costaz C Bernot A Nicaud S Jaffe D Fisher S Lutfalla G Dossat C Segurens B Dasilva C Salanoubat M Levy M Boudet N Castellano S Anthouard V Jubin C Castelli V Katinka M Vacherie B Biemont C Skalli Z Cattolico L Poulain J De Berardinis V Cruaud C Duprat S Brottier P Coutanceau JP Gouzy J Parra G Lardier G Chapple C McKernan KJ McEwan P Bosak S Kellis M Volff JN Guigo R Zody MC Mesirov J Lindblad-Toh K Birren B Nusbaum C Kahn D Robinson-Rechavi M Laudet V Schachter V Quetier F Saurin W Scarpelli C Wincker P Lander ES Weissenbach J Roest Crollius H Genome duplication in the teleost fish Tetraodon nigroviridis reveals the early vertebrate proto-karyotype Nature 2004 431 946 957 15496914 10.1038/nature03025
Dickinson JL Norris BJ Jensen PH Antalis TM The C-D interhelical domain of the serpin plasminogen activator inhibitor-type 2 is required for protection from TNF-alpha induced apoptosis Cell Death Differ 1998 5 163 171 10200461 10.1038/sj.cdd.4400324
Chuang TL Schleef RR Identification of a nuclear targeting domain in the insertion between helices C and D in protease inhibitor-10 J Biol Chem 1999 274 11194 11198 10196205 10.1074/jbc.274.16.11194
Wilczynska M Lobov S Ohlsson PI Ny T A redox-sensitive loop regulates plasminogen activator inhibitor type 2 (PAI-2) polymerization Embo J 2003 22 1753 1761 12682008 10.1093/emboj/cdg178
Lobov S Wilczynska M Bergstrom F Johansson LB Ny T Structural bases of the redox-dependent conformational switch in the serpin PAI-2 J Mol Biol 2004 344 1359 1368 15561148 10.1016/j.jmb.2004.10.010
Jensen PH Schuler E Woodrow G Richardson M Goss N Hojrup P Petersen TE Rasmussen LK A unique interhelical insertion in plasminogen activator inhibitor-2 contains three glutamines, Gln83, Gln84, Gln86, essential for transglutaminase-mediated cross-linking J Biol Chem 1994 269 15394 15398 7910824
Ritchie H Lawrie LC Crombie PW Mosesson MW Booth NA Cross-linking of plasminogen activator inhibitor 2 and alpha 2-antiplasmin to fibrin(ogen) J Biol Chem 2000 275 24915 24920 10816585 10.1074/jbc.M002901200
Darnell GA Antalis TM Johnstone RW Stringer BW Ogbourne SM Harrich D Suhrbier A Inhibition of retinoblastoma protein degradation by interaction with the serpin plasminogen activator inhibitor 2 via a novel consensus motif Mol Cell Biol 2003 23 6520 6532 12944478 10.1128/MCB.23.18.6520-6532.2003
Ragg H Lokot T Kamp PB Atchley WR Dress A Vertebrate serpins: construction of a conflict-free phylogeny by combining exon-intron and diagnostic site analyses Mol Biol Evol 2001 18 577 584 11264410
Hillier LW Miller W Birney E Warren W Hardison RC Ponting CP Bork P Burt DW Groenen MA Delany ME Dodgson JB Chinwalla AT Cliften PF Clifton SW Delehaunty KD Fronick C Fulton RS Graves TA Kremitzki C Layman D Magrini V McPherson JD Miner TL Minx P Nash WE Nhan MN Nelson JO Oddy LG Pohl CS Randall-Maher J Smith SM Wallis JW Yang SP Romanov MN Rondelli CM Paton B Smith J Morrice D Daniels L Tempest HG Robertson L Masabanda JS Griffin DK Vignal A Fillon V Jacobbson L Kerje S Andersson L Crooijmans RP Aerts J van der Poel JJ Ellegren H Caldwell RB Hubbard SJ Grafham DV Kierzek AM McLaren SR Overton IM Arakawa H Beattie KJ Bezzubov Y Boardman PE Bonfield JK Croning MD Davies RM Francis MD Humphray SJ Scott CE Taylor RG Tickle C Brown WR Rogers J Buerstedde JM Wilson SA Stubbs L Ovcharenko I Gordon L Lucas S Miller MM Inoko H Shiina T Kaufman J Salomonsen J Skjoedt K Wong GK Wang J Liu B Yu J Yang H Nefedov M Koriabine M Dejong PJ Goodstadt L Webber C Dickens NJ Letunic I Suyama M Torrents D von Mering C Zdobnov EM Makova K Nekrutenko A Elnitski L Eswara P King DC Yang S Tyekucheva S Radakrishnan A Harris RS Chiaromonte F Taylor J He J Rijnkels M Griffiths-Jones S Ureta-Vidal A Hoffman MM Severin J Searle SM Law AS Speed D Waddington D Cheng Z Tuzun E Eichler E Bao Z Flicek P Shteynberg DD Brent MR Bye JM Huckle EJ Chatterji S Dewey C Pachter L Kouranov A Mourelatos Z Hatzigeorgiou AG Paterson AH Ivarie R Brandstrom M Axelsson E Backstrom N Berlin S Webster MT Pourquie O Reymond A Ucla C Antonarakis SE Long M Emerson JJ Betran E Dupanloup I Kaessmann H Hinrichs AS Bejerano G Furey TS Harte RA Raney B Siepel A Kent WJ Haussler D Eyras E Castelo R Abril JF Castellano S Camara F Parra G Guigo R Bourque G Tesler G Pevzner PA Smit A Fulton LA Mardis ER Wilson RK Sequence and comparative analysis of the chicken genome provide unique perspectives on vertebrate evolution Nature 2004 432 695 716 15592404 10.1038/nature03154
Lander ES Linton LM Birren B Nusbaum C Zody MC Baldwin J Devon K Dewar K Doyle M FitzHugh W Funke R Gage D Harris K Heaford A Howland J Kann L Lehoczky J LeVine R McEwan P McKernan K Meldrim J Mesirov JP Miranda C Morris W Naylor J Raymond C Rosetti M Santos R Sheridan A Sougnez C Stange-Thomann N Stojanovic N Subramanian A Wyman D Rogers J Sulston J Ainscough R Beck S Bentley D Burton J Clee C Carter N Coulson A Deadman R Deloukas P Dunham A Dunham I Durbin R French L Grafham D Gregory S Hubbard T Humphray S Hunt A Jones M Lloyd C McMurray A Matthews L Mercer S Milne S Mullikin JC Mungall A Plumb R Ross M Shownkeen R Sims S Waterston RH Wilson RK Hillier LW McPherson JD Marra MA Mardis ER Fulton LA Chinwalla AT Pepin KH Gish WR Chissoe SL Wendl MC Delehaunty KD Miner TL Delehaunty A Kramer JB Cook LL Fulton RS Johnson DL Minx PJ Clifton SW Hawkins T Branscomb E Predki P Richardson P Wenning S Slezak T Doggett N Cheng JF Olsen A Lucas S Elkin C Uberbacher E Frazier M Gibbs RA Muzny DM Scherer SE Bouck JB Sodergren EJ Worley KC Rives CM Gorrell JH Metzker ML Naylor SL Kucherlapati RS Nelson DL Weinstock GM Sakaki Y Fujiyama A Hattori M Yada T Toyoda A Itoh T Kawagoe C Watanabe H Totoki Y Taylor T Weissenbach J Heilig R Saurin W Artiguenave F Brottier P Bruls T Pelletier E Robert C Wincker P Smith DR Doucette-Stamm L Rubenfield M Weinstock K Lee HM Dubois J Rosenthal A Platzer M Nyakatura G Taudien S Rump A Yang H Yu J Wang J Huang G Gu J Hood L Rowen L Madan A Qin S Davis RW Federspiel NA Abola AP Proctor MJ Myers RM Schmutz J Dickson M Grimwood J Cox DR Olson MV Kaul R Shimizu N Kawasaki K Minoshima S Evans GA Athanasiou M Schultz R Roe BA Chen F Pan H Ramser J Lehrach H Reinhardt R McCombie WR de la Bastide M Dedhia N Blocker H Hornischer K Nordsiek G Agarwala R Aravind L Bailey JA Bateman A Batzoglou S Birney E Bork P Brown DG Burge CB Cerutti L Chen HC Church D Clamp M Copley RR Doerks T Eddy SR Eichler EE Furey TS Galagan J Gilbert JG Harmon C Hayashizaki Y Haussler D Hermjakob H Hokamp K Jang W Johnson LS Jones TA Kasif S Kaspryzk A Kennedy S Kent WJ Kitts P Koonin EV Korf I Kulp D Lancet D Lowe TM McLysaght A Mikkelsen T Moran JV Mulder N Pollara VJ Ponting CP Schuler G Schultz J Slater G Smit AF Stupka E Szustakowski J Thierry-Mieg D Thierry-Mieg J Wagner L Wallis J Wheeler R Williams A Wolf YI Wolfe KH Yang SP Yeh RF Collins F Guyer MS Peterson J Felsenfeld A Wetterstrand KA Patrinos A Morgan MJ de Jong P Catanese JJ Osoegawa K Shizuya H Choi S Chen YJ Initial sequencing and analysis of the human genome Nature 2001 409 860 921 11237011 10.1038/35057062
Aparicio S Chapman J Stupka E Putnam N Chia JM Dehal P Christoffels A Rash S Hoon S Smit A Gelpke MD Roach J Oh T Ho IY Wong M Detter C Verhoef F Predki P Tay A Lucas S Richardson P Smith SF Clark MS Edwards YJ Doggett N Zharkikh A Tavtigian SV Pruss D Barnstead M Evans C Baden H Powell J Glusman G Rowen L Hood L Tan YH Elgar G Hawkins T Venkatesh B Rokhsar D Brenner S Whole-genome shotgun assembly and analysis of the genome of Fugu rubripes Science 2002 297 1301 1310 12142439 10.1126/science.1072104
The Chimpanzee Sequencing and Analysis Consortium Initial sequence of the chimpanzee genome and comparison with the human genome Nature 2005 437 69 87 16136131 10.1038/nature04072
Altschul SF Gish W Miller W Myers EW Lipman DJ Basic local alignment search tool J Mol Biol 1990 215 403 410 2231712 10.1006/jmbi.1990.9999
Burge C Karlin S Prediction of complete gene structures in human genomic DNA J Mol Biol 1997 268 78 94 9149143 10.1006/jmbi.1997.0951
Thompson JD Higgins DG Gibson TJ CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice Nucleic Acids Res 1994 22 4673 4680 7984417
Page RD TreeView: an application to display phylogenetic trees on personal computers Comput Appl Biosci 1996 12 357 358 8902363
|
16305753
|
PMC1308813
|
CC BY
|
2021-01-04 16:32:47
|
no
|
BMC Genomics. 2005 Nov 23; 6:167
|
utf-8
|
BMC Genomics
| 2,005 |
10.1186/1471-2164-6-167
|
oa_comm
|
==== Front
BMC Health Serv ResBMC Health Services Research1472-6963BioMed Central London 1472-6963-5-721628197710.1186/1472-6963-5-72Research ArticleCost-consciousness among Swiss doctors: a cross-sectional survey Bovier Patrick A [email protected] Diane P [email protected] Thomas V [email protected] Department of community medicine, Geneva University Hospitals, 24 Micheli-du-Crest, CH-1211 Geneva 14, Switzerland2 Quality of Care Unit, Geneva University Hospitals, 24 Micheli-du-Crest, CH-1211 Geneva 14, Switzerland3 Department of Health Services, Box 357660, University of Washington, Seattle, WA 98195, USA2005 10 11 2005 5 72 72 17 2 2005 10 11 2005 Copyright © 2005 Bovier et al; licensee BioMed Central Ltd.2005Bovier et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Knowing what influences physicians attitudes toward health care costs is an important matter, because most health care expenditures are the results of doctors' decisions. Many decisions regarding medical tests and treatments are influenced by factors other than the expected benefit to the patient, including the doctor's demographic characteristics and concerns about cost and income.
Methods
Doctors (n = 1184) in Geneva, Switzerland, answered questions about their cost-consciousness, practice patterns (medical specialty, public.vs. private sector, number of patients per week, time spent with a new patient), work satisfaction, and stress from uncertainty. General linear models were used to identify independent risk factors of higher cost-consciousness.
Results
Most doctors agreed that trying to contain costs was their responsibility ("agree" or "totally agree": 90%) and that they should take a more prominent role in limiting the use of unnecessary tests (92%); most disagreed that doctors are too busy to worry about costs (69%) and that the cost of health care is only important if the patient has to pay for it out-of-pocket (88%). In multivariate analyses, cost-consciousness was higher among doctors in the public sector, those who saw fewer patients per week, who were most tolerant of uncertainty, and who were most satisfied with their work.
Conclusion
Thus even in a setting with very high health care expenditures, doctors' stated cost-consciousness appeared to be generally high, even though it was not uniformly distributed among them.
==== Body
Background
Because most health care expenditures are the results of doctors' decisions, whether doctors are cost-conscious is an important matter. Many decisions regarding medical tests and treatments are influenced by factors other than the expected benefit to the patient, including the doctor's demographic characteristics [1,2], training [3-6], work context [7,8], financial incentives [9,10] and information about costs [11-13]. Medical decisions are also influenced by subjective considerations, including risk aversion [14,15], tolerance for uncertainty [16], and concerns about cost and income [17].
In this paper, we are interested in cost-consciousness, defined as a concern to contain costs of health care borne by society [18]. Cost-consciousness was the first factor identified from a series of different attitudes hypothesized to influence physicians' resource use [19], based on a conceptual model of medical care process [20] and a review of physician decision making [21]. Goold et al. found that cost-consciousness was associated with lower self-reported estimates of resource use. However, whether cost-consciousness leads to less costly medical care is not known. While those who pay for health care may regard this evolution as desirable, cost-consciousness creates a possible conflict of interest. According to traditional medical deontology, the sole concern of a doctor should be the welfare of the patient, regardless of costs to society [22]. The tensions caused by such conflicts of interest are particularly acute in managed care settings [23,24]. Thus doctors may have good reasons not to be cost-conscious.
The aims of this study were to assess the cost-consciousness of doctors practicing in public and private settings in the Swiss canton Geneva, and to identify practice patterns and doctors' characteristics associated with cost-consciousness.
Methods
Setting
Switzerland has the second highest health care expenditures per capita in the world, behind the United States, and devotes about 11% of the GNP to health care [25,26]. Doctors in private practice who are paid on a fee-for-service schedule provide most ambulatory care, and most hospitals are public, subsidised by local governments, where doctors are salaried (in addition, some senior hospital doctors have private practice privileges). To achieve social solidarity, the Health Insurance Law makes compulsory the purchase by households of a fairly comprehensive package of health benefits, which includes ambulatory treatment, inpatient care, home nursing care, and some health promotion activities. The basic health insurance coverage can be contracted from approximately 90 private insurance carriers, which are not allowed to earn profits from the mandated benefit package. Insurers can offer different schemes for health care provision such as patients' free choice of physician ("any-willing-provider" or compulsory contracting) or preferred providers' contracts (general practitioner-gatekeeper model or restricted network of providers). Restriction on the choice of the provider results in lower premiums for the patient. According to official statistics, in 2001, about 45% of the insured chose the lowest permissible deductible and only 9% the highest; about 8% chose a managed care policy, mainly the general practitioner-gatekeeper model. Regarding the organisation of care, the Swiss health care system has therefore been qualified as a "regulated competition system without managed care" [27]. The canton Geneva has the highest health care costs and medical density in the country. In 1998, there were 58 doctors per 10000 residents in canton Geneva – 31 in private practice, and 27 in public hospitals.
Study design
We conducted a cross-sectional mail survey of doctors practicing in canton Geneva, Switzerland, during the fall of 1998. Doctors were identified from membership files of the Geneva Medical Association (1370 members) and the Swiss Association of Interns/Registrars, Geneva Section (906 members). After exclusion of 54 duplicate records, 10 pre-test participants, 97 doctors who had incorrect addresses and 121 who did not practice clinical medicine, 1994 doctors were eligible.
Measurement of cost-consciousness
Attitude toward costs (cost-consciousness) was measured with a validated 6-item instrument [19]. Two items explore the doctors' opinion about health care costs in general, three items explore their attitudes regarding the costs of tests and procedures, and a last item probes the importance of out-of-pocket payments (Table 1). The items were scored on a five-point Likert scale anchored by 1: totally disagree and 5: totally agree. Negatively worded items were reversed so that a higher score would mean greater cost-consciousness. We translated the instrument into French using a standardized procedure (three independent translations, selection of a consensus translation by an expert panel, pre-tests among 10 physicians) (Appendix 1). The internal consistency of the translated scale was similar to that of the original scale (Cronbach α : 0.68, versus 0.74 for original) and a factor analysis indicated that the translated items represented a single dimension. We computed a summary cost-consciousness score whenever at least 3 of the 6 items were present.
Table 1 Original English and French translation of the cost-consciousness scale.
Trying to contain costs is the responsibility of every physician. Essayer de maîtriser les coûts de la santé est la responsabilité de chaque médecin.
There is currently too much emphasis on costs of tests and procedures. * Actuellement, on met trop l'accent sur le coût des examens et des interventions médicales. *
Doctors need to take a more prominent role in limiting use of unnecessary tests. Les médecins doivent jouer un rôle plus important dans les efforts pour limiter l'emploi d'examens inutiles.
Doctors are too busy to worry about the costs of tests and procedures. * Les médecins sont trop occupés pour se soucier du coût des examens et des interventions médicales. *
The cost of a test or medication is only important if the patient has to pay for it out-of-pocket. * Le coût d'un examen ou d'un traitement est important seulement si le patient doit le payer de sa poche. *
It is unfair to ask physicians to be cost-conscious and still keep the welfare of their patients foremost in their minds. * Il est injuste de demander aux médecins de penser à réduire les coûts de la santé, tout en se préoccupant avant tout du bien-être de leurs patients. *
Answer scale : 1 : Totally disagree/Pas du tout d'accord – 5 : Totally agree/Tout à fait d'accord *: reversed score for the summary scale
Predictors of cost-consciousness
Determinants of cost-consciousness included socio-demographic characteristics (age, sex), time since graduation from medical school, type of practice (public versus private), medical specialty, workload characteristics (number of patients per week, time spent with a new patient), stress from uncertainty, and work-related satisfaction.
Stress from uncertainty refers to physicians' reaction to their limitations of professional knowledge, problems of diagnosis, ambiguities of treatment and outcome, unpredictability of patients response, and variations in physicians' attitudes, values and perceptions of risk [28-33]. Items of a stress from uncertainty scale [28] were translated into French using the same procedure as for the cost-consciousness scale. Factor analysis confirmed that the items represented a single dimension. The internal scale consistency coefficient was 0.88.
Work-related satisfaction was measured by a 17-item instrument based on the work of the Society of General Internal Medicine Career Satisfaction Study Group [34], containing questions addressing satisfaction with workload, intellectual stimulation, leisure time, work stress, relations with patients, autonomy in treating patients, autonomy in specialist referrals, overall quality of care, relations with peers, relations with nurses and other non medical staff, administrative burden, continuing medical education, enjoyment of work, respect and prestige, payment mechanism, current income, and job satisfaction overall [35]. The internal scale consistency coefficient of a work related satisfaction scale including all 17 items was 0.88.
Data analysis
Descriptive statistics were calculated for each item of the cost-consciousness instrument (mean, standard deviation, floor & ceiling effect). Summary scores were calculated by averaging the responses to the relevant items, whenever half or more were present. For comparisons with the original [19], the cost-consciousness score was rescaled between 0 and 24; to facilitate the interpretation of regression models, we also standardized this score to mean 50 and standard deviation 10.
We explored the relationships of the cost-consciousness scale with socio-demographic and job characteristics of the respondents, stress from uncertainty, and work related satisfaction. We used analysis of variance, including tests for linear trend where indicated, to test these associations. Because units of all scales are arbitrary, we categorized the continuous predictors into quartiles, because correlation or regression coefficients can be difficult to interpret, while quartiles are more intuitive. Furthermore, this approach also allows the reader to check whether the association is linear or not. Multivariate models were used to adjust for potential confounders, identify independent risk factor of higher cost-consciousness score, and compute adjusted means. All statistical tests were two-tailed, with a significance level of 0.05.
Results
After the first mailing and two reminders, 1184 doctors (59%) responded to the survey. Two thirds were men (784, 66%) and their mean age was 44.9 years (SD: 10.9). Most respondents were in private practice (748, 63%), 362 (31%) were hospital interns or registrars in the public sector, and 67 (6%) held senior posts in the public hospitals (7 missing data). Respondents included 402 primary care doctors (generalists, general internists: 34%), 196 internal medicine specialists (17%), 83 paediatricians (7%), 178 psychiatrists (15%), and 325 (27%) other specialists (surgical specialties, radiologists). On average, they had graduated from medical school 17.8 years earlier (SD: 10.3).
Doctors' attitude toward health care costs
The distribution of the answers to most items was skewed toward the higher end of the response scale (Table 2). Almost all doctors (90%) agreed that trying to contain costs was their responsibility, and that they should take a more prominent role in limiting the use of unnecessary tests (92%). A majority of respondents (69%) disagreed with statements that doctors are too busy to worry about costs, and that the cost of a medication is only important if the patient has to pay for it out-of-pocket (88%). Only one out of four doctors (27%) thought that it is unfair to ask doctors to be cost-conscious and still keep the welfare of their patients foremost in their minds. Almost two-thirds (61%) agreed that there is currently too much emphasis on the costs of tests and procedures. The cost-consciousness score could be computed for 1176 respondents (99%); its mean was 16.8, SD 3.6, and quartiles 13.8, 16.8, and 19.2 when scaled between 0 and 24 (21). Its ditribution was close to normal.
Table 2 Descriptive statistics of the items of the cost-consciousness scale of 1184 doctors in Geneva, Switzerland, 1998.
Percent responding
Item Missing Totally disagree Disagree Not sure Agree Totally agree Mean (SD)
Trying to contain costs is the responsibility of every doctor. 0.6 1.5 2.7 5.8 45.8 44.2 4.3 (0.8)
There is currently too much emphasis on costs of tests and procedures. a 1.4 5.2 18.4 15.8 42.2 18.3 2.5 (1.1)
Doctors need to take a more prominent role in limiting use of unnecessary tests. 0.9 0.7 1.4 6.2 50.6 41.1 4.3 (0.7)
Doctors are too busy to worry about the costs of tests and procedures. a 0.8 33.6 35.0 16.2 12.3 2.9 3.8 (1.1)
The cost of a test or medication is only important if the patient has to pay for it out-of-pocket. a 0.9 59.7 28.7 6.6 3.8 1.3 4.4 (0.9)
It is unfair to ask doctors to be cost-conscious and still keep the welfare of their patients foremost in their minds. a 1.4 19.6 34.7 18.8 19.0 8.0 3.4 (1.2)
a: Negatively worded items were reversed so that a higher score would mean greater cost-consciousness.
Relationship to socio-demographic and work-related characteristics
Cost-consciousness was similar in men and women (Table 3), but younger doctors had slightly higher scores. Doctors in the public sector were in general more cost-conscious, in particular the senior staff across all age categories (results not shown). Internal medicine specialists and paediatricians had higher cost-consciousness scores, while doctors in surgical and technical specialties had the lowest (results not shown). Doctors who saw fewer patients per week and spend more time with each new patient were also more cost-conscious. Cost-consciousness scores were also higher for doctors who reported lower stress from uncertainty and higher work-related satisfaction (Table 3).
Table 3 Relationships of cost-consciousness to socio-demographic and work-related characteristics of 1184 Swiss doctors, Geneva in Switzerland, 1998.
N % Cost-consciousness T score P value
Sex 0.68a
Men 784 66 50.1
Women 400 34 49.8
Age (years) (21 missing) 0.05a
<35 263 23 50.4
35–50 568 49 50.5
>50 332 28 48.9
Years since graduation from medical school (17 missing) 0.22a
0–10 325 28 50.5
11–17 290 25 49.7
18–24 272 23 50.6
>24 280 24 49.1
Type of practice 0.001a
Public sector, in training 368 31 50.9
Public sector, senior staff 68 6 53.5
Private sector 748 63 49.3
Self-reported number of patients per week (83 missing) <0.001b
<26 269 24 51.6
26–50 409 37 50.1
51–75 195 18 49.1
>75 228 21 48.6
Self-reported time spent (minutes) with a new patient (66 missing) 0.001b
<31 382 34 48.6
31–45 324 29 50.2
46–60 344 31 50.4
>60 68 6 52.5
Stress from uncertainty (20 missing) <0.001b
Lowest quartile (1 – 2.13) 269 23 51.8
2nd quartile (2.15 – 2.67) 305 26 50.6
3rd quartile (2.69 – 3.23) 336 29 50.3
Highest quartile (3.25 – 5.0) 254 22 46.9
Work-related satisfaction (9 missing) <0.001b
Lowest quartile (1.0 – 4.5) 284 24 48.2
2nd quartile (4.53 – 5.0) 338 29 49.6
3rd quartile (5.06 – 5.47) 271 23 50.7
Highest quartile (5.5 – 7.0) 282 24 51.6
a: ANOVA, difference between groups
b: ANOVA, test for linearity.
Multivariate analyses
Multivariate analysis identified the following independent risk factors for higher cost-consciousness: type of practice (doctors in public sectors had higher scores), lower workload (number of patients per week), lower stress from uncertainty, and higher work-related satisfaction (Table 4). Other factors that were significant in the univariate analyses, such as specialty and self-reported time with a new patient, were not anymore significant after adjustment for type of practice, self-reported number of patient seen per week, stress from uncertainty and work related satisfaction. No sociodemographic characteristics were statistically associated with cost-consciousness in the multivariate analysis. We also explored the possible interactions between these factors, but none were significant.
Table 4 Multivariate predictors of cost-consciousness in 1184 Swiss doctors, Geneva, Switzerland, 1998.
Adjusted mean cost consciousness 95% CI
P value
Type of practice 0.005a
Public sector, in training 51.2 50.0 to 52.4
Public sector, senior staff 51.8 49.3 to 54.3
Private sector 49.0 48.2 to 49.8
Self-reported number of patients per week 0.002b
<26 51.9 50.6 to 53.2
26–50 51.2 50.0 to 52.4
51–75 50.3 48.7 to 52.0
>75 49.1 47.5 to 50.7
Stress from uncertainty <0.001b
Lowest quartile (1 – 2.13) 52.4 51.0 to 53.9
2nd quartile (2.15 – 2.67) 51.4 50.1 to 52.8
3rd quartile (2.69 – 3.23) 50.9 49.5 to 52.2
Highest quartile (3.25 – 5.0) 47.9 46.4 to 49.4
Work-related satisfaction <0.001b
Lowest quartile (1.0 – 4.5) 48.5 47.0 to 49.9
2nd quartile (4.53 – 5.0) 50.3 48.9 to 51.6
3rd quartile (5.06 – 5.47) 51.3 49.9 to 52.8
Highest quartile (5.5 – 7.0) 52.5 51.0 to 54.0
a: ANOVA, difference between groups
b: ANOVA, test for linearity.
Discussion
In a setting where medical density and health care expenditures are particularly high, we found that a majority of doctors agreed that trying to contain costs was their responsibility, that they should worry about the costs of tests and procedures they order, that they should take a more prominent role in limiting the use of unnecessary tests, and that the cost of a test or medication is not only important if the patient has to pay for it out-of-pocket. In multivariate analyses, cost-consciousness was higher among doctors in the public sector, those who saw fewer patients a week, those who were most tolerant of uncertainty, and those most satisfied with their work.
The absolute levels of cost consciousness were similar in our sample and in the previous report from an academic medical centre in the United States [19]: on a scale of 0 to 24, the mean was 16.8 (SD 3.6) in Geneva, Switzerland, and 17.2 (SD 3.1) in Ann Arbor, Michigan. Whether attitudes toward cost-containment are different in other settings would require other similar surveys.
Practice patterns
The difference between the private and the public sectors may have several origins. It is possible that working in a public hospital raises the doctors' sensitivity to health care costs. Alternatively, the public sector may attract doctors who are more cost-conscious than those who choose the private sector. This issue could be resolved by following doctors prospectively, as they move from the public hospital into private practice.
Regardless of the setting, doctors who saw fewer patients per week were also more cost-conscious. In a fee-for-service system, a doctor's income is directly related to the number of patients seen. Doctors who see more patients per week may be more concerned about their income and less by costs incurred by society. However, the relationship was similar among salaried hospital doctors, who have no such financial incentive. It is possible that hospital doctors who see fewer patients may have academic or administrative roles that cause them to be more concerned about health care costs.
Stress from uncertainty
The association between cost-consciousness and higher tolerance for uncertainty is compatible with a previous report which linked higher anxiety due to uncertainty with higher health care expenditures in a Medicare health maintenance organisation [16]. Thus there is evidence that higher stress from uncertainty is associated with higher expenditures and, as we have found, lower cost-consciousness. From this, it follows that lower cost-consciousness and higher expenditures could be associated. To our knowledge, no study has yet studied this relationship.
Work-related satisfaction
The association with work-related satisfaction was substantial. This is of interest as work-related satisfaction correlates with several health outcomes, in particular with the quality of care [36,37]. Although this association may be causal in either direction, we believe that work-related satisfaction may influence positively cost-consciousness: doctors who are happy about their work may think more about societal issues than unhappy doctors. Unlike other correlates of cost-consciousness, keeping doctors satisfied with their work is a modifiable factor. Nevertheless, whether improvements in doctor satisfaction translate into greater cost-consciousness remains to be seen.
Strengths and limitations
The main strength of this study is its population-based sampling frame; the results are fairly representative of the attitudes of physicians in Geneva, Switzerland. The response rate of 59% raises the issue of selection bias, but is comparable to other surveys that depend on the cooperation of busy practitioners [2,38,39]. We could not assess response bias in depth, because the two medical organisations gave us only the adresses of their members, with no other socio-demographic information. The only caracteristic that we could track was sex, and its distribution was similar among those who responded and those who did not.
The main limitation of the study was its cross-sectional design, which precludes any formal conclusion about the causality of the associations between cost-consciousness and its correlates. Second, we were not able to obtain direct measures of doctor "costliness", such as the mean cost per patient, and thus cannot establish whether cost-consciousness influences actual costs of patient care. Third, we do not know to what extent respondents gave socially desirable answers about this rather sensitive topic. Finally, as the scales we used are not perfectly reliable, random misclassification may have caused the relationships we observed to appear weaker than they are in reality.
Conclusion
In this study conducted in a setting where health care expenditures are among the highest in the world, doctors appeared to be generally concerned about the need to contain costs. Therefore failure to control costs does not seem to stem from a general lack of concern about costs among doctors. Nevertheless, levels of cost-consciousness were variable. Doctors in private practice, who saw the most patients, who were the least tolerant with uncertainty, and the least satisfied with their work were also the least cost-conscious.
Competing interests
All three authors have no conflict of interest. Financial support for this study was provided entirely by a grant from the Swiss National Science Foundation (3200-053377). The funding agreement ensured the authors' independence in designing the study, interpreting the data, writing, and publishing the report.
Authors' contributions
PAB participated in the design of the study, collected the data, performed the statistical analysis and wrote the initial manuscript. DPM participated in the design of the study and reviewed the manuscript. TVP participated in the design of the study, analysis of the data and reviewed the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors wish to thank all members of the Geneva Medical Association and of the Swiss Association of Interns/Registrars, Geneva Section, who participated to the survey, their respective chairmen, Dr Claude Aubert and Dr Marco Bettoni who supported this work.
==== Refs
Woodward CA Hutchison B Norman GR Brown JA Abelson J What factors influence primary care physicians' charges for their services? An exploratory study using standardized patients CMAJ 1998 158 197 202 9469140
Halm EA Causino N Blumenthal D Is gatekeeping better than traditional care? A survey of physicians' attitudes JAMA 1997 278 1677 1681 9388087 10.1001/jama.278.20.1677
Glassman PA Kravitz RL Petersen LP Rolph JE Differences in clinical decision making between internists and cardiologists Arch Intern Med 1997 157 506 512 9066454 10.1001/archinte.157.5.506
Selby JV Grumbach K Quesenberry CPJ Schmittdiel JA Truman AF Differences in resource use and costs of primary care in a large HMO according to physician specialty Health Serv Res 1999 34 503 518 10357287
Spann SJ Interspeciality differences in medical resource utilization J Fam Pract 2000 49 18 19 10678334
Skootsky SA Slavin S Wilkes MS Attitudes toward managed care and cost containment among primary care trainees at 3 training sites Am J Manag Care 1999 5 1397 1404 10662413
Reichert S Simon T Halm EA Physicians' attitudes about prescribing and knowledge of the costs of common medications Arch Intern Med 2000 160 2799 2803 11025790 10.1001/archinte.160.18.2799
Kerr EA Mittman BS Hays RD Siu AL Leake B Brook RH Managed care and capitation in California: how do physicians at financial risk control their own utilization? Ann Intern Med 1995 123 500 504 7661493
Hemenway D Killen A Cashman SB Parks CL Bicknell WJ Physicians' responses to financial incentives. Evidence from a for-profit ambulatory care center N Engl J Med 1990 322 1059 1063 2320066
Lindenthal JJ Lako CJ van der Waal MA Tymstra T Andela M Schneider M Quality and cost of healthcare: a cross-national comparison of American and Dutch attitudes Am J Manag Care 1999 5 173 181 10346513
Tierney WM Miller ME McDonald CJ The effect on test ordering of informing physicians of the charges for outpatient diagnostic tests N Engl J Med 1990 322 1499 1504 2186274
Lin YC Miller SR The impact of price labeling of muscle relaxants on cost consciousness among anesthesiologists J Clin Anesth 1998 10 401 403 9702621 10.1016/S0952-8180(98)00053-1
Korn LM Reichert S Simon T Halm EA Improving physicians' knowledge of the costs of common medications and willingness to consider costs when prescribing J Gen Intern Med 2003 18 31 37 12534761 10.1046/j.1525-1497.2003.20115.x
Fiscella K Franks P Zwanziger J Mooney C Sorbero M Williams GC Risk aversion and costs: a comparison of family physicians and general internists J Fam Pract 2000 49 12 17 10691394
Zaat JO van Eijk JT General practitioners' uncertainty, risk preference, and use of laboratory tests Med Care 1992 30 846 854 1518316
Allison JJ Kiefe CI Cook EF Gerrity MS Orav EJ Centor R The association of physician attitudes about uncertainty and risk taking with resource use in a Medicare HMO Med Decis Making 1998 18 320 329 9679997
Hemenway D Health Care Prices & Choices 1993 Lanham, Maryland, University Press of America
Greene HL Goldberg RJ Beattie H Russo AR Ellison RC Dalen JE Physician attitudes toward cost containment. The missing piece of the puzzle Arch Intern Med 1989 149 1966 1968 2774777 10.1001/archinte.149.9.1966
Goold SD Hofer T Zimmerman M Hayward RA Measuring physician attitudes toward cost, uncertainty, malpractice, and utilization review J Gen Intern Med 1994 9 544 549 7823224
Donabedian A Aspects of Medical Care Administration: Specifying Requirements for Health Care 1973 Cambridge, MA, Harvard University Press
Eisenberg JM Doctors' Decisions and the Costs of Medical Care 1986 Ann Arbor, MI, Health Administration Press
Feldman DS Novack DH Gracely E Effects of managed care on physician-patient relationships, quality of care, and the ethical practice of medicine: a physician survey Arch Intern Med 1998 158 1626 1632 9701096 10.1001/archinte.158.15.1626
Pellegrino ED Interests, obligations, and justice: some notes toward an ethic of managed care J Clin Ethics 1995 6 312 317 8750589
Pellegrino ED Allocation of resources at the bedside: the intersections of economics, law, and ethics Kennedy Inst Ethics J 1994 4 309 317 10138751
Frei A Hunsche E The Swiss health care system Eur J Health Economics 2001 2 76 78
European Observatory on Health Care Systems Health care systems in transition: Switzerland 2000 2003
Reinhardt UE The Swiss health system: regulated competition without managed care JAMA 2004 292 1227 1231 15353536 10.1001/jama.292.10.1227
Kassirer JP Our stubborn quest for diagnostic certainty. A cause of excessive testing N Engl J Med 1989 320 1489 1491 2497349
Geller G Tambor ES Chase GA Holtzman NA Measuring physicians' tolerance for ambiguity and its relationship to their reported practices regarding genetic testing Med Care 1993 31 989 1001 8231339
Gerrity MS DeVellis RF Earp JA Physicians' reactions to uncertainty in patient care: a new measure and new insights Medical Care 1990 28 724 736 2385142
Pearson SD Goldman L Orav EJ Guadagnoli E Garcia TB Johnson PA Lee TH Triage decisions for emergency department patients with chest pain: do physicians' risk attitudes make the difference? J Gen Intern Med 1995 10 557 564 8576772
Poses RM De Saintonge DM McClish DK Smith WR Huber EC Clemo FL Schmitt BP Alexander-Forti D Racht EM Colenda CC Centor RM An international comparison of physicians' judgments of outcome rates of cardiac procedures and attitudes toward risk, uncertainty, justifiability, and regret Med Decis Making 1998 18 131 140 9566446
Friedmann PD Brett AS Mayo-Smith MF Differences in generalists' and cardiologists' perceptions of cardiovascular risk and the outcomes of preventive therapy in cardiovascular disease Ann Intern Med 1996 124 414 421 8554250
Konrad TR Williams ES Linzer M McMurray J Pathman DE Gerrity M Schwartz MD Scheckler WE Van Kirk J Rhodes E Douglas J Measuring physician job satisfaction in a changing workplace and a challenging environment. SGIM Career Satisfaction Study Group. Society of General Internal Medicine Med Care 1999 37 1174 1182 10549620 10.1097/00005650-199911000-00010
Bovier PA Perneger TV Predictors of work satisfaction among physicians Eur J Public Health 2003 13 299 305 14703315 10.1093/eurpub/13.4.299
Grol R Mokkink H Smits A van Eijk J Beek M Mesker P Mesker-Niesten J Work satisfaction of general practitioners and the quality of patient care Fam Pract 1985 2 128 135 4043602
Haas JS Cook EF Puopolo AL Burstin HR Cleary PD Brennan TA Is the professional satisfaction of general internists associated with patient satisfaction? J Gen Intern Med 2000 15 122 128 10672116 10.1046/j.1525-1497.2000.02219.x
Ginsburg ME Kravitz RL Sandberg WA A survey of physician attitudes and practices concerning cost-effectiveness in patient care West J Med 2000 173 390 394 11112752 10.1136/ewjm.173.6.390
Franks P Williams GC Zwanziger J Mooney C Sorbero M Why do physicians vary so widely in their referral rates? J Gen Intern Med 2000 15 163 168 10718896 10.1046/j.1525-1497.2000.04079.x
|
16281977
|
PMC1308814
|
CC BY
|
2021-01-04 16:31:49
|
no
|
BMC Health Serv Res. 2005 Nov 10; 5:72
|
utf-8
|
BMC Health Serv Res
| 2,005 |
10.1186/1472-6963-5-72
|
oa_comm
|
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-1061630768210.1186/1471-2334-5-106Case ReportFailure of levofloxacin treatment in community-acquired pneumococcal pneumonia Endimiani Andrea [email protected] Gioconda [email protected] Alessia A [email protected] Francesco [email protected] Paolo [email protected] Antonio Q [email protected] Laboratory of Microbiology and Virology, University of Insubria and Ospedale di Circolo e Fondazione Macchi, Varese, Italy2 Department of Infectious Diseases, University of Insubria and Ospedale di Circolo e Fondazione Macchi, Varese, Italy2005 24 11 2005 5 106 106 29 7 2005 24 11 2005 Copyright © 2005 Endimiani et al; licensee BioMed Central Ltd.2005Endimiani et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Streptococcus pneumoniae is the leading cause of community-acquired pneumonia (CAP). High global incidence of macrolide and penicillin resistance has been reported, whereas fluoroquinolone resistance is uncommon. Current guidelines for suspected CAP in patients with co-morbidity factors and recent antibiotic therapy recommend initial empiric therapy using one fluoroquinolone or one macrolide associated to other drugs (amoxicillin, amoxicillin/clavulanate, broad-spectrum cephalosporins). Resistance to fluoroquinolones is determined by efflux mechanisms and/or mutations in the parC and parE genes coding for topoisomerase IV and/or gyrA and gyrB genes coding for DNA gyrase. No clinical cases due to fluoroquinolone-resistant S. pneumoniae strains have been yet reported from Italy.
Case presentation
A 72-year-old patient with long history of chronic obstructive pulmonary disease and multiple fluoroquinolone treatments for recurrent lower respiratory tract infections developed fever, increased sputum production, and dyspnea. He was treated with oral levofloxacin (500 mg bid). Three days later, because of acute respiratory insufficiency, the patient was hospitalized. Levofloxacin treatment was supplemented with piperacillin/tazobactam. Microbiological tests detected a S. pneumoniae strain intermediate to penicillin (MIC, 1 mg/L) and resistant to macrolides (MIC >256 mg/L) and fluoroquinolones (MIC >32 mg/L). Point mutations were detected in gyrA (Ser81-Phe), parE (Ile460-Val), and parC gene (Ser79-Phe; Lys137-Asn). Complete clinical response followed treatment with piperacillin/tazobactam.
Conclusion
This is the first Italian case of community-acquired pneumonia due to a fluoroquinolone-resistant S. pneumoniae isolate where treatment failure of levofloxacin was documented. Molecular analysis showed a group of mutations that have not yet been reported from Italy and has been detected only twice in Europe. Treatment with piperacillin/tazobactam appears an effective means to inhibit fluoroquinolone-resistant strains of S. pneumoniae causing community-acquired pneumonia in seriously ill patients.
==== Body
Background
Streptococcus pneumoniae is the leading cause of community-acquired pneumonia (CAP) and a major cause of meningitis and otitis media. Recent reports show an high global incidence of macrolide and penicillin resistance, whereas fluoroquinolone (FQ) resistance is not frequent [1-3]. Current guidelines for suspected bacterial CAP in patients with co-morbidity factors and recent antibiotic therapy recommend initial empiric therapy using one respiratory FQ or one macrolide associated to other drugs (amoxicillin, amoxicillin/clavulanate, broad-spectrum cephalosporins) [4,5].
Resistance to FQ in S. pneumoniae is determined by efflux mechanisms and/or mutations in the quinolone resistance-determining regions (QRDRs) of parC and parE genes coding for topoisomerase IV and/or gyrA and gyrB genes coding for DNA gyrase [6]. The most frequent resistance mechanism detected in S. pneumoniae is represented by associated mutations of both gyrA and parC genes [1,3,7]. A single parC mutation appears to represent the first step conferring low-level FQ resistance [6]. A second mutational step in the gyrA gene produces high-level resistance to FQ [6]. Both mutations are probably selected in response to the widespread use of FQ [8,9]. Ciprofloxacin and levofloxacin use is commonly associated with parC gene mutations, whereas use of other FQ appears to favor the selection of mutations in other genes [6]. Here, we report the first Italian case of CAP in which failure of levofloxacin treatment was associated to high-level FQ resistance in S. pneumoniae.
Case presentation
Microbiological methods
Bacterial identification (ID) and antimicrobial susceptibility testing (AST) were achieved with the Phoenix System using SMIC/ID-2 panels (Becton Dickinson Diagnostic Systems, Sparks, MD). Specie ID was confirmed using optochin and bile solubility tests (Taxo Disc, BBL, Becton Dickinson) [10]. Determinations of minimal inhibitory concentration (MIC) were obtained with the Etest method (AB Biodisk, Solna, Sweden) on Mueller-Hinton agar with 5% sheep blood (Oxoid, Milano, Italy). MIC values were interpreted according to current criteria of the Clinical and Laboratory Standards Institute [11].
Amplification of gyrA, gyrB,parC, parE, and pneumolysin genes were performed by PCR using reported oligonucleotides [12-14]. Direct DNA sequencing was obtained using the ABI Prism Big Dye terminator Cycle Sequencing Kit and the ABI Prism 310 sequencer (Applied Biosystems, Foster City, CA). DNA sequences were compared with sequences present in the GenBank database (gyrA, accession No. AB010387; gyrB, accession No. Z67740; parE and parC, accession No. Z67739). Efflux mechanisms were investigated by broth dilution in Mueller-Hinton medium containing 5% sheep blood using serial dilution of ciprofloxacin in the presence or in the absence of 10 mg/L reserpine (Sigma-Aldrich, St. Louis, MO). A fourfold or greater decrease of the ciprofloxacin MIC in the presence of reserpine was considered as indicator of efflux mechanisms [15].
Clinical data and results
In March 2005, a 72-year-old patient with long history of chronic obstructive pulmonary disease (COPD), congenital bullous emphysema, and aortic valve incompetence developed abrupt-onset fever, cough, increased sputum production, and dyspnea. Over the past few years, solely on the basis of clinical data, the patient had received multiple treatments with oral levofloxacin and ciprofloxacin for recurrent lower respiratory tract infections (LRTi). At the onset, the family physician prescribed oral levofloxacin (500 mg, bid) and paracetamol (500 mg, tid). Three days later, because of non-remitting fever and acute respiratory insufficiency, the patient was admitted to the Emergency Room of the Ospedale di Circolo e Fondazione Macchi (Varese, Italy). On the basis of physical examination and hypoxemia (pO2 = 45.4 mm Hg) the patient was intubated. A central venous catheter was implanted. Blood analysis showed the following results: erythrocytes 4.23 × 106/μL, hemoglobin 13.3 g/dL, hematocrit 38.8%, platelets 201 × 103/μL, total leukocytes 15.75 × 103/μL (neutrophils 78.2%, lymphocytes 12.1%), ESR (erythrocyte sedimentation rate) 86.2 mm/h, CRP (C-reactive protein) 118 mg/L, creatinine 1.02 mg/dL, BUN (blood urea nitrogen) 26 mg/dL. Chest radiograph showed accentuated peribronchial interstitium and hilar expansion. Intravenous corticosteroids and bronchodilators were given. Twelve hours later, the patient was transferred to the Intensive Care Unit (ICU) where 4 bronchoalveolar lavages (BAL) were obtained. Levofloxacin was continued by the i.v. route (500 mg, bid) supplemented with piperacillin/tazobactam (2.25 g tid). Serial BAL were also obtained during the first 3 days of hospitalization.
Microbiological tests detected a multidrug-resistant (MDR) S. pneumoniae isolate (106 colony forming units/mL) in the four BAL samples obtained at admission. Isolates had the following MIC values: penicillin G (1 mg/L), piperacillin (1 mg/L), piperacillin/tazobactam (1 mg/L), ceftriaxone (0.5 mg/L), cefotaxime (0.5 mg/L), cefepime (0.5 mg/L), imipenem (0.064 mg/L), erythromycin (>256 mg/L), claritromicyn (>256 mg/L), clindamycin (>256 mg/L mg/L), levofloxacin (>32 mg/L), ciprofloxacin (>32 mg/L), moxifloxacin (>32 mg/L), tetracycline (0.5 mg/L), chloramphenicol (16 mg/L), trimethoprim-sulfamethoxazole (4 mg/L), rifampin (0.5 mg/L) vancomycin (0.25 mg/L), teicoplanin (0.125 mg/L), linezolid (0.75 mg/L), and telithromycin (≤0.006 mg/L).
By direct DNA sequencing, point mutations in the gyrA (Ser81-Phe) and parE (Ile460-Val) genes were detected in all isolates. Two point mutations of the parC gene (Ser79-Phe; Lys137-Asn) were also present. No mutations were detected in the gyrB gene. The reserpine response assay failed to evidence FQ efflux. The pneumolysin gene was detected in all isolates.
Prompt clinical response followed the empirical treatment given at the ICU. On day-two, the patient was afebrile, hypoxemia resolved (pO2 = 157 mm Hg), inflammatory markers were decreased (total leukocytes 7.19 × 103/μL, neutrophils 76.1%, lymphocytes 12.0%, ESR 45 mm/h, CRP 57 mg/L), and blood cultures were negative. BAL samples obtained on day-2 and at subsequent times were negative. Five days after hospital admission, levofloxacin was suspended and piperacillin/tazobactam was reduced (1.25 g tid). Respiratory assistance was stopped on day-12. Antibiotic treatment ended on day-14. Complete recovery ensued and the patient was discharged on day-16.
Conclusion
The impact of antibiotic resistance on the treatment outcome of patients with CAP is a matter of discussion [4]. FQ are often the primary choice for empirical treatment of patients with co-morbidity factors such as COPD, diabetes, and renal or congestive heart failure [4,5]. A few reports show an association between FQ resistance and treatment failure in CAP caused by S. pneumoniae [16,17]. No cases have been yet reported from Italy, possibly due to the low incidence of FQ-resistant strains [3].
The present report describes the first Italian case of CAP due to a FQ-resistant strain of S. pneumoniae. Failure of oral levofloxacin was observed. According to current criteria [11], the isolate was MDR i.e., intermediate to penicillin, resistant to chloramphenicol, trimethoprim-sulfamethoxazole, macrolides, clindamycin and FQ. The strain showed single mutations of gyrA and parE genes associated with two point-mutations of the parC gene. High-level resistance to FQ resulted (MIC >32 mg/L). This justified the initial treatment failure when the patient was given levofloxacin alone. Molecular analysis showed a group of mutations that have not yet been reported from Italy [3,18]. Association of the reported mutations in gyrA, parE and parC genes has been detected only twice in Europe [3], but is relatively more frequent in the USA [7]. As demonstrated by studies of S. pneumoniae strains exposed in vitro to FQ, ciprofloxacin and levofloxacin use is commonly associated with parC gene mutations, whereas use of other FQ appears to favor the selection of mutations in other genes [6].
The reported clinical case suggests that the following precautions need to be used for patients with specific LRTi risk factors: i) before therapy, respiratory samples should be obtained to identify the responsible agents and to define antimicrobial susceptibility [16]; ii) patients previously treated with FQ may be given these drugs, but need strict clinical monitoring during the first 3 days of therapy to evaluate the clinical response [19]; iii) since only 1% of FQ-resistant S. pneumoniae strains are also resistant to broad-spectrum beta-lactams, seriously ill patients can be safely treated with combinations of FQ plus broad-spectrum beta-lactams [20]. Empiric treatment of CAP in patients seriously compromised using piperacillin/tazobactam (as in the reported case) appears an effective means to inhibit FQ-resistant strains of S. pneumoniae. In particular, piperacillin/tazobactam appears indicated for ICU patients since, in contrast to monotherapy with extended-spectrum cephalosporins (e.g., ceftriaxone, cefotaxime), does not favor the selection of ESBL-positive enterobacteria and/or chromosomal beta-lactamase hyperproducers (e.g., Pseudomonas aeruginosa) [21,22]. Additionally, this drug's spectrum is wider than that of 3rd–4th generation cephalosporins.
In Italy FQ consumption is higher than in the rest of Europe [8]. Thus, it seems possible that over the next years an increased prevalence of FQ-resistant S. pneumoniae strains will be observed. For this reason the detection of a community-acquired MDR isolate of S. pneumoniae is of special concern. Emerging resistance traits in this species underline the need of in vitro tests based on MIC data in order to select the most appropriate drugs for preventing the dissemination of epidemic clones.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
AE and GB performed microbiological tests, analyzed the data and prepared the manuscript.
AAB carried out molecular assays.
PG diagnosed, cured and provided clinical data of the patient.
FL and AQT coordinated the study and helped writing the manuscript.
All authors read and approved the final version of the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
We wish to thank Vito Elia, and Silvano Colella for their expert technical assistance. Written consent was obtained from the patient for publication of the study. This work was supported by a PRIN grant from the Italian Ministry for Education, University, and Research (MIUR, Rome, Italy). AE is a Ph.D. student of the University of Insubria, Varese, Italy.
==== Refs
Canton R Morosini M Enright MC Morrissey I Worldwide incidence, molecular epidemiology and mutations implicated in fluoroquinolone-resistant Streptococcus pneumoniae : data from the global PROTEKT surveillance programme J Antimicrobial Chemother 2003 52 944 952 10.1093/jac/dkg465
Brown SD Farrell DJ Morrissey I Prevalence and molecular analysis of macrolide and fluoroquinolone resistance among isolates of Streptococcus pneumoniae collected during the 2000–2001 PROTEKT US Study J Clin Microbiol 2004 42 4980 4987 15528684 10.1128/JCM.42.11.4980-4987.2004
Reinert RR Reinert S van der Linden M Cil MY Al-Lahham A Appelbaum P Antimicrobial susceptibility of Streptococcus pneumoniae in eight European countries from 2001 to 2003 Antimicrob Agents Chemother 2005 49 2903 2913 15980367 10.1128/AAC.49.7.2903-2913.2005
Mandell LA Bartlett JG Dowell SF File TM JrMusher DM Whitney C Update of practice guidelines for the management of community-acquired pneumonia in immunocompetent adults Clin Infect Dis 2003 37 1405 1433 14614663 10.1086/380488
Gilbert DN Moellering RC JrEliopoulos GM Sande MA The Sanford guide to antimicrobial therapy 2005 Thirty-fifth Hyde Park, VT, USA
Eliopoulos GM Quinolone resistance mechanisms in pneumococci Clin Infect Dis 2004 350 356 10.1086/382692
Pletz MWR McGee L Jorgensen J Beall B Facklam RR Whitney CG Klugman KP the Active Bacterial Core Surveillance Team Levofloxacin-resistant invasive Streptococcus pneumoniae in the United States: evidence for clonal spread and the impact of conjugate pneumococcal vaccine Antimicrob Agents Chemother 2004 48 3491 3497 15328116 10.1128/AAC.48.9.3491-3497.2004
Goossens H Ferech M Vander Stichele R Elseviers M for the ESAC Project Group Outpatient antibiotic use in Europe and association with resistance: a cross-national database study Lancet 2005 365 579 587 15708101
Pérez-Trallero E Marimón JM González A Ercibengoa M Larruskain J In vivo development of high-level fluoroquinolone resistance in Streptococcus pneumoniae in chronic obstructive pulmonary disease Clin Infect Dis 2005 41 560 564 16028169 10.1086/432062
Arbique JC Poyart C Trieu-Cuot P Quesne G da Glória S Carvalho M Steigerwalt AG Morey RE Jackson D Davidson RJ Facklam RR Accuracy of phenotypic and genotypic testing for identification of Streptococcus pneumoniae and description of Streptococcus pseudopneumoniae sp. nov J Clin Microbiol 2004 42 4686 4696 15472328 10.1128/JCM.42.10.4686-4696.2004
Clinical and Laboratory Standards Institute Performance standards for antimicrobial susceptibility testing 2005
Janoir C Zeller V Kitzis M-D Moreau NJ Gutmann L High-level fluoroquinolone resistance in Streptococcus pneumoniae requires mutations in parC and gyrA Antimicrob Agents Chemother 1996 40 2760 2764 9124836
Perichon B Tankovic J Courvalin P Characterization of a mutation in the parE gene that confers fluoroquinolone resistance in Streptococcus pneumoniae Antimicrob Agents Chemother 1997 41 1166 1167 9145891
Dagan R Shriker O Hazan I Leibovitz E Greenberg D Schlaeffer F Levy R Prospective study to determined clinical relevance of detection of pneumococcal DNA in sera of children by PCR J Clin Microbiol 1998 36 669 673 9508293
Bast DJ Low DE Duncan CL Kilburn L Mandell LA Davidson RJ de Azavedo JCS Fluoroquinolone resistance in clinical isolates of Streptococcus pneumoniae : contributions of type II topoisomerase mutations and efflux to levels of resistance Antimicrob Agents Chemother 2000 44 3049 3054 11036021 10.1128/AAC.44.11.3049-3054.2000
Davidson R Cavalcanti R Brunton JL Bast DJ de Azavedo JCS Kibsey P Fleming C MLT Low DE Resistance to levofloxacin and failure of treatment of pneumococcal pneumonia N Engl J Med 2002 346 747 750 11882730 10.1056/NEJMoa012122
Fuller JD Low DE A review of Streptococcus pneumoniae infection treatment failures associated with fluoroquinolone resistance Clin Infect Dis 2005 41 118 121 15937772 10.1086/430829
Montanari MP Tili E Cochetti I Mingoia M Manzin A Varaldo PE Molecular characterization of clinical Streptococcus pneumoniae isolates with reduced susceptibility to fluoroquinolones emerging in Italy Microb Drug Resist 2004 10 209 217 15383164
Anderson KB Tan JS File TM JrDiPersio JR Willey BM Low DE Emergence of levofloxacin-resistant pneumococci in immunocompromised adults after therapy for community-acquired pneumonia Clin Infect Dis 2003 37 376 381 12884162 10.1086/376642
Jones RN Fritsche TR Sader HS Therapeutic options among broad-spectrum β-lactam for infections caused by levofloxacin-nonsusceptible Streptococcus pneumoniae Diagn Microbiol Infect Dis 2005 52 129 133 15964501 10.1016/j.diagmicrobio.2004.12.009
Kollef MH Optimizing antibiotic therapy in the intensive care unit setting Critical Care 2001 5 189 195 11511331 10.1186/cc1022
Graffunder EM Preston KE Evans AM Venezia RA Risk factors associated with extended-spectrum β-lactamase-producing organisms at a tertiary care hospital J Antimicrob Chemother 2005 56 139 145 15917283 10.1093/jac/dki180
|
16307682
|
PMC1308815
|
CC BY
|
2021-01-04 16:28:16
|
no
|
BMC Infect Dis. 2005 Nov 24; 5:106
|
utf-8
|
BMC Infect Dis
| 2,005 |
10.1186/1471-2334-5-106
|
oa_comm
|
==== Front
BMC Int Health Hum RightsBMC International Health and Human Rights1472-698XBioMed Central London 1472-698X-5-71628007910.1186/1472-698X-5-7Research ArticlePrevalence of mental disorders and torture among Tibetan refugees: A systematic review Mills Edward J [email protected] Sonal [email protected] Timothy H [email protected] Robert M [email protected] Sonam [email protected] Joanna [email protected] James J [email protected] Dept. of Clinical Epidemiology & Biostatistics, McMaster University, Hamilton, Canada2 Department of Medicine, Wake Forest University, Winston-Salem, USA3 Rollins School of Public Health, Emory University, Atlanta, USA4 Department of Community Health Sciences, University of Manitoba, Winnipeg, Canada5 Department of Political Science, York University, Toronto, Canada6 Department of Psychiatry and Peace Studies, McMaster University, Hamilton, Canada7 St. Michael's Hospital, University of Toronto, Toronto, Canada2005 9 11 2005 5 7 7 30 6 2005 9 11 2005 Copyright © 2005 Mills et al; licensee BioMed Central Ltd.2005Mills et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Many Tibetan refugees flee Tibet in order to escape physical and mental hardships, and to access the freedoms to practice their culture and religion. We aimed to determine the prevalence of mental illnesses within the refugee population and determine the prevalence of previous torture reported within this population.
Methods
We performed a systematic literature search of 10 electronic databases from inception to May 2005. In addition, we searched the internet, contacted all authors of located studies, and contacted the Tibetan Government-in-exile, to locate unpublished studies. We included any study reporting on prevalence of mental illness within the Tibetan refugee populations. We determined study quality according to validation, translation, and interview administration. We calculated proportions with exact confidence intervals.
Results
Five studies that met our inclusion criteria (total n = 410). All studies were conducted in North India and 4 were specifically in adult populations. Four studies provided details on the prevalence of torture and previous imprisonment within the populations. The prevalence of post-traumatic stress disorder ranged from 11–23%, anxiety ranged from 25–77%, and major depression ranged from 11.5–57%.
Conclusion
Our review indicates that the prevalence of serious mental health disorders within this population is elevated. The reported incidence of torture and imprisonment is a possible contributor to the illnesses. Non-government organizations and international communities should be aware of the human rights abuses being levied upon this vulnerable population and the mental health outcomes that may be associated with it.
==== Body
Background
In 1950, China began its occupation of Tibet. Since then, many Tibetans have fled to Nepal and India. More than 50 years of occupation have been accompanied by forced population displacements, widespread hunger, restrictions on cultural and religious freedoms, well-documented political violence against specific cultural groups, mass arrests, imprisonment of political prisoners and execution. Human rights groups have documented at least 60 deaths of peaceful demonstrators since 1987. [1] The current Tibetan population of Tibet is estimated at 6 million, with an undetermined number of Chinese occupants. This repression against Tibetans has resulted in large refugee populations in neighboring countries. It is estimated that more than 150,000 Tibetan refugees reside in the neighboring countries of Bhutan, Nepal, and India;[1] a generous token from such poor countries.
Due to political oppression, cultural oppression, and a desire to practice their religion, a growing number of Tibetans seek to escape to Tibetan settlements in India, the seat of the Tibetan Government-in-exile. The most common route is through Nepal. Depending on the point of departure and the type of transport used, the journey from Tibet to Nepal can take several days to months1. Most refugees cross the high mountains along commonly used escape routes to access Nepal. The Tibetan Refugee Transit Centre (TRTC), established by the United Nations High Commissioner for Refugees (UNHCR) in Katmandu, Nepal estimate that an average of 2,500 Tibetan refugees arrive into Nepal every year, with an equal number unsuccessful in their journey due to death or captivity. The TRTC assists refugees with their onward journey to India. The social background of those Tibetans who manage to flee has remained fairly constant over the past few decades, with a majority proportion of new arrivals being minors, monks and nuns. Nomads, farmers, and unemployed persons make up the remaining proportion. The number of attempted Tibetan refugee flights increases or decreases depending on the vigilance of Chinese border patrols.
Globally, at the start of 2004, there are an estimated 9.7 million refugees, included in a total of just over 17 million listed as 'persons of concern' for the UNHCR (including asylum seekers, internally displaced persons, returned refugees still being monitored, stateless persons, and refugees)[2]. Mental disorders are often overlooked in refugee populations. A recent analysis of mental illnesses within refugees residing in developed countries found that the prevalence of post-traumatic stress disorder (PTSD) was roughly 1 in 10, and major depressive disorder was roughly 1 in 20[3]. We previously examined the experiences of recent refugees in the flight from Tibet to Nepal and discovered that those interviewed were at a greater susceptibility for mental health issues than expected, due to traumatic events, torture, and unfamiliarity with their new surroundings[4]. The goal of this analysis was to review the prevalence of mental illnesses reported amongst the Tibetan refugee populations, and secondarily examine the reported incidences of torture by the Tibetan refugees. To determine this, we conducted a systematic review of the available literature.
Methods
Inclusion/exclusion criteria
Eligible studies assessed the prevalence of mental health disorders among Tibetan refugees in Nepal and India, and repatriated refugees in Tibet. Studies had to report original communications on the assessment of mental health outcomes using a measurement tool. We excluded qualitative studies, case reports and studies addressing the political experiences of participants. We additionally excluded studies assessing the mental health of Chinese nationals residing in Tibet.
With the aid of an information specialist, we searched the following 10 data bases (from inception to May 2005): AMED, CINAHL, Cochrane CENTRAL and the Cochrane library, EMBASE, ERIC, MedLine via PubMed, HEALTHSTAR, Psych-info, Sociological abstracts, and Web of Science. We additionally searched the internet using Google as well as bibliographies of relevant papers. Where appropriate we searched using the single term Tibet* as the yield was manageable. We contacted the authors of all studies for clarification of study outcomes, garnering a response rate of 100%. The Tibetan Government in Exile was contacted to determine if they were aware of unpublished research.
We used the definition of torture as stated in the United Nations Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment and Punishment (UN Torture Convention)[5]. Torture, according to this convention, is defined as any act by which severe pain or suffering, whether physical or mental, is intentionally inflicted on a person for such purposes as obtaining from him or a third person information or a confession, punishing him for an act he or a third person has committed or is suspected of having committed, or intimidating or coercing him or a third person, or for any reason based on discrimination of any kind, when such pain or suffering is inflicted by or at the instigation of or with the consent or acquiescence of a public official or other person acting in an official capacity[5].
Two reviewers worked independently and in duplicate, to review the abstracts and full text versions of identified reports and to adjudicate their inclusion.
Data abstraction
Two reviewers, working independently extracted data from the included studies using a standardized form. We specifically abstracted information on the following: population; duration of time outside of Tibet; number of children, clergy, elderly, and women; access to mental health services; and mental health outcomes (PTSD, anxiety and depression) identified using standardized measurement tools. We additionally extracted information on the number reporting imprisonment and torture.
Methodological quality
In order to determine the validity of the interventions used to assess mental health status, we examined if these measurement tools had been translated and validated in the Tibetan language. We additionally assessed whether the tool was provided by interviewer administration or self-administered. We assessed whether reliability had been assessed and whether authors evaluated construct validity.
We determined the diagnostic criteria of the screening tools, where available, as follows: Harvard Trauma Questionnaire, scores above 2.5 are considered strongly suspicious of PTSD;[6] Hopkins Symptom Checklist-25, scores above 1.75 on the individual (anxiety and depression) of the HSCL-25 is consistent with significant emotional distress and correlates with the presence of diagnosable psychiatric morbidity. [7]
Statistical analysis
Where proportions of populations were provided, we calculated the exact confidence intervals around the proportions. We did not pool results due to the heterogeneity of populations and methodologies employed. We tested interrater reliability using the K statistic. All statistics were performed using StatsDirect (Manchester, 2003).
Results
Our systematic searches yielded 21 relevant abstracts. Thirteen were excluded as review articles. Of the remainder, 8 were selected for further examination and 4 were excluded as they were either qualitative or examined physical health outcomes. Four studies were included from academic publications[1,8-10]. One additional study was included from the non-government Organization (NGO) Physicians for Human Rights[11]. K for inclusion was ≥ 1, indicating perfect agreement. Table 1 describes the study populations. All studies were conducted in North India and all studies met the requirements of reporting specific methodological issues.
Table 1 Name, year Setting;
Country Population Design N Duration outside
Tibet Torture
survivors (%) Detained (n) Measurement
tools Outcomes
PHR, 1997 Dharamsala Refugee Reception Center,
Transit School for Young Adults,
and a Buddhist monastery; India Mixed population
including clergy Cross-sectional 258 (191 men
and 67 women) Median 6 months 21% 20% 10-item validated
questionnaire to assess
torture (UN Torture Convention) 15% (10–19%)
Torture survivors 53 21% HSCL-25-anxiety-depression 53% (40–66%)
40% (28–53%)
DSM-IV criteria
for PTSD 23% (13–39%)
Holtz, 1998 35 tortured Tibetan nuns
and students with a cohort
of 35 closely matched subjects. Retrospective cohort 70 Mean of
30.9 months
(SD = 19.2) 50% For torture survivors
HSCL-25
- anxiety
- depression
54% (44–63%)
14% (8–22%)
Servan-Schreiber, 1998 Tibetan Childrens Village,
Dharamsala, India Children Cross-sectional 61 Mean 13.3 months NA NA DSM-IV PTSD
DSM-IV Major depression 11.5 (3.5–19.5%)
11.5 (3.5–19.5%)
DSM-IV Major depression 11.5 (3.5–19.5%)
Crescenzi, 2002 Confidential setting,
Dharamsala, India Purposeful adults
-76 previously imprisoned
74 never imprisoned Case-control 150 0–4 years 95% 50% For total sample
HSCL-25
- anxiety
- depression
HTQ assessment
63% (55–70%)
57% (49–65%)
20 (14–27%)
Terheggen, 2001 Refugee camp,
North India Random selection using
random sequencing Cross-sectional 76 1 month-2.5 years 12% 25% PTI
HSCL-25
- anxiety
- depression Median 3
(range 0–3)
25% (17–36%)
42% (32–53%)
HSCL, Hopkins Symptom Check List, PTI, Post Traumatic Inventory
Three studies reported on PTSD (total n = 264) [1,8,11]. The prevalence of post-traumatic stress disorder ranged from 11–23%. Four studies reported on prevalence of an anxiety disorder (total n = 349) and ranged from 25–77% [8-11]. Five studies reported on prevalence of major depressive disorder (total n = 357)[1,8-10] and ranged from 11.5–57%.
Below we report on the individual studies and include details on the study population; incidence of torture, and; prevalence of mental disorders, respectively.
Physicians for Human Rights (PHR)
Study population
Physicians for Human Rights[11] (PHR) conducted a large convenience sample survey using a validated checklist to assess the prevalence of torture and imprisonment among newly arrived Tibetan refugees in Dharamsala, India. Fifty-five individuals reported being tortured (21%). Those who were reportedly tortured tended to be young at their first episode of torture (mean age 19.5, range 13–28). Fifty-eight percent (32/55) were less than 21 years old, and 15% (8/55) were 16 years old or younger at the time of their torture.
Incidence of torture
Sixty percent (33/55) of the torture victims in this study reported being subjected to three or more different forms of torture in addition to threats or verbal abuse. Torture by electric shock commonly employed the use of cattle prods, including applying the cattle prods to the genitals, mouth and eyes. One individual reported being immersed in a tub of water, before being forced to lay down on an electrified metal bed. Two forms of torture that PHR identified, which previously have been infrequently reported, included being forced to stare at the sun for prolonged periods of time, and having blood drawn against the individual's will.
Prevalence of mental disorders
PHR further examined mental health outcomes within the torture survivors subgroup by utilizing the Hopkins Symptom Checklist (HSCL-25) to assess emotional distress and the DSM-IV criteria to diagnose PTSD. Seventy-eight percent of the torture survivors to whom the Hopkins Symptoms Checklist-25 was administered were suffering from significant symptoms of anxiety and/or depression. Eighty-eight percent of the torture survivors surveyed concerning PTSD symptoms reported recurrent, intrusive memories, including flashbacks and nightmares of their abuse. Twenty-three percent of these individuals met DSM-IV criteria for PTSD.
Holtz
Study population
A retrospective cohort study was conducted in Dharamsala India in 1995[9]. Thirty five refugee Tibetan nuns and lay students who were arrested and tortured in Tibet were matched with 35 Tibetan refugee controls not previously arrested or tortured. The groups were similar in most respects, with the exception that the tortured group had been more politically active than the controls, whilst living in Tibet. The mean age of the torture survivors was 25.8 years(SD = 2.3 years).
Incidence of torture
The torture events took place during a mean of 21 months of captivity. The mean length of abuse was 38 days. Fifty-seven percent (20) reported solitary confinement, with a mean confinement period of 5.4 weeks. Overall, the survivors reported exposure to a mean of 13.3 different forms (SD = 3.2) of torture, ranging from 7 to 21 types. Only 43% (15) of those tortured were ever charged with a crime, and fewer (7%) actually brought to a trial. One survivor remembered the presence of a doctor while being tortured. Forms of torture characteristic of Chinese prisons in Tibet include electrical shocks to the body (86%), forced standing (86%), exposure to bright sunshine (69%), and having blood extracted without consent (50%).
Prevalence of mental disorders
Participants were administered the HSCL-25, to evaluate anxiety symptoms, affective disturbances, somatic complaints, and social impairment. The prevalence of symptom scores in the clinical range for both cohorts was 41.4% for anxiety symptoms and 14.3% for depressive symptoms. The torture survivors had a statistically significant higher proportion of elevated anxiety scores than did the nontortured cohort(54.3% vs. 28.6%, p =.01). Depressive scores were not different between groups. The results suggest that torture has long-term consequences on mental health over and above the effects of being uprooted, fleeing one's country, and living in exile.
Servan-Schreiber
Study population
A cross-sectional study examining refugee children in Dharamsala India[1], was conducted to asses PTSD and depression within a Tibetan children's population. Sixty one randomly selected children from the Tibetan Children's Village School were administered a questionnaire by a psychiatrist assessing DSM-IV diagnosis for PTSD and major depression. The average children's age was 12.6 (SD 2.5) and was equally representative of males to females (30 vs. 31).
Prevalence of mental disorders
PTSD was diagnosed in 11.5% of the children (95% CI, 3.5–19.5%), with a further 18% with suspected PTSD (95% CI, 8.4–27.4%). There was a trend for more cases of full criteria PTSD amongst children who had recently arrived (<18 months) (25% vs 6.7%, chi-square, P = 0.06). Major depression was diagnosed in 11.5% of the population with 18.9% of all children over the age of 13 meeting the full criteria for major depression. The study did not report the proportions or quantifiable number of children who had experienced torture or imprisonment, but did include children who had been physically traumatized.
Crescenzi et al
Study population
In a case-control study in Dharamsala, India[8], Crescenzi et al. compared mental health disorders among 76 previously imprisoned refugees to 74 never imprisoned refugees. Forty-five percent of the total sample had been clergy in Tibet. All participants were >18 years of age at the time of escape from Tibet. Utilizing a detailed translation of the HSCL-25 checklist, the authors assess anxiety and depression, as well as listed traumatic events.
Incidence of torture
Although the previously imprisoned group had a higher incidence of torture, most participants in both groups had experienced some degree of torture. The most common torture techniques across both groups were beatings (73%), electrical torture (43%), being forced to provide blood (19%), and being kept naked (25%). In addition to torture, both groups reported traumatic events which may affect mental health. These included sleep deprivation (36%), witnessing murder (37%), kidnapping of family and friends (37%), and disappearances of family and friends (13%).
Prevalence of mental disorders
The authors found that refugees who were previously imprisoned were more likely to experience mental health issues than those never imprisoned, although both groups had elevated mental health symptoms [HSCL-anxiety scores 22.5 (SD = 6.5) vs. 18.7 (SD = 6.8), P = 0.001] [HSCL-depressive scores 28.5 (SD 7.4) vs. 27.6 (SD 7.4), P=>0.05]. PTSD was diagnosed in 20% of the total sample and anxiety and depression were diagnosable amongst 63% and 57% of the total sample, respectively.
Terheggen et al
Study population
In a cross-sectional study of refugee students residing at a North India refugee camp[10], Terheggen et al. interviewed 75 individuals about past traumatic events and assessed the correlation between traumatic events and severe mental health diagnosis. Participants were new refugees and ranged in age from 18–29 years. Sixteen percent made weekly visits to their physician.
Incidence of torture
The issues identified by the participants as most traumatic included: destruction of religious signs (43%), relative or friend tortured (28%), relative or friend disappeared (26%), relative or friend sent to prison (28%), being tortured oneself (12%), being sent to prison (25%), and fearing for one's own life (32%). In addition, important traumatic events were identified that were unexpected, such as: being forbidden to speak own language (3%), being forbidden to live according to one's own religion (15%), and being forbidden to live according to one's own culture (9%).
Prevalence of mental disorders
The study found a high prevalence of anxiety (25%) and depression (42%) among the population. Importantly, the study found a correlation between incidences of trauma and mental health diagnoses (0.67). This study is the only study which identified the trauma that the refugees perceived when witnessing the destruction of the religious artifacts.
Conclusion
The findings of this systematic review should be of interest to NGOs, policy makers and politicians. We found that the prevalence of mental illnesses within the Tibetan refugee population is remarkably high, with increased incidences in the many refugees reporting severe torture and violations of human rights. Governments and NGOs should be aware that human rights violations are impacting mental health outcomes in this vulnerable population.
There are certain limitations to interpreting our study. We performed a systematic review using standard searching techniques. It is possible however, that NGO reports were not published and as such, were excluded without our knowledge. However, we did consult the Tibetan Government in Exile for their knowledge regarding additional studies and none were located. A further limitation is that not all studies examined the impact of torture on mental health illnesses, and it is possible, and likely, that torture victims have increased susceptibility to mental illness, such as PTSD. Studies used diverse instruments with which to determine the prevalence of mental health illnesses. In the studies we included, 4 used screening tools to evaluate anxiety and depressive disorders [8-11] Only 1 consistently used diagnostic tools.[1] The accurate assessment of psychiatric disorders is difficult to ensure in epidemiological studies, especially in non-Western populations, for whom the validity of measures developed in Western populations may be restricted[3,12-14]. In our review, most assessments of anxiety and depression were determined by the Hopkins Symptom Check List (HSCL), a tool with extensive use internationally[15]. The HSCL-25 is the most commonly used scale to assess anxiety and depressive symptoms amongst refugee populations. Translations of the HSCL-25 have been validated against clinically-assessed diagnoses of anxiety and major depression[7], for use in several south Asian refugee groups[7,16,17], but have not tested for construct validity in a Tibetan population. Recent translations of the HSCL-25 into a variety of languages have demonstrated good psychometric properties and have been widely used in adolescents and adults to assess psychiatric morbidity in traumatized populations and refugee groups worldwide[12,18-20] The test-retest reliability of this instrument has been found to be r = .90 for Southeast Asian refugees[21]. The diagnosis of PTSD was however, varied. Two studies used DSM-IV diagnostic case tools[1,11], while 1 used the Post-Traumatic Inventory (PTI)[10] and 1 used the Harvard Trauma Questionnaire[8], a well-validated questionnaire with good correlation with clinical diagnosis. The studies examined were all conducted in North India, after the refugees had arrived from Nepal. It may be that the incidence of mental health illnesses are higher during their stay in Nepal and decrease with time, as was observed in several included studies[1,8,9]. Further, the studies included in our review were generally limited in methodological quality. No study assessed test-retest reliability or construct validity. Only one study used a random-sampling technique[10] All studies were limited in their sample size and it is possible that this contributes to a sampling bias leading to a higher incidence.[22] A final limitation is that this study only examined incidents of torture reported by the refugees. Other Tibetans who are 'persons of concern' for the UNHCR are also likely to have experienced torture, and were not included for analysis in this review.
Our findings indicate that the prevalence of mental health illnesses within this population are higher than those reported in most refugee populations[3,22]. Of particular concern is the high prevalence of PTSD and depression among the children examined[1]. The Tibetan Government in Exile should be urged to encourage children to report on violations and provide specialized treatment programmes for children. These findings are consistent with the findings of a previous qualitative study indicating that children are often witnesses and victims of human rights violations and exploitation, including sexual exploitation[4].
Our findings demonstrate that torture is commonly reported amongst Tibetan refugees, and that those who have experienced torture often suffer significant psychological effects. The findings in this study have a number of important implications for the health of the Tibetan refugee community and the international community as well. The Tibetan Government in Exile provides a specialized torture treatment program only for political torture survivors [23]. Access to care for all survivors of torture, both political and non-political, needs to be provided, as well as specialized programs for children. In addition, this study illustrates the paucity of generalizable information about the mental health of this population in what appears to be a sustained assault on the health and dignity of the Tibetan culture and community. In our review, we found that severe torture, including electric cattle prods on genitals and oral cavities, as well as forced blood draws, was routinely reported. In addition, many refugees cited their mental health illnesses as stemming from witnessing their family and friends murdered.
Under the Criminal Law of the People's Republic of China, it is forbidden to extort a confession by torture[24]. Furthermore, China is a signatory to a number of international human rights conventions, legally binding under international law, including the Convention Against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment[25], and the Convention on the Rights of the Child[26]. Based upon our systematic review, wherein all systematic torture and detainment occurred in China, China is in clear violation of these conventions. [11]
All governments and other economic partners to China should be aware that human rights violations are unacceptable, and should pressure the Chinese government for adherence to international human rights and humanitarian standards, including prohibitions against the use of torture. In order to prevent further illness and human rights violations, international NGOs should bring pressure to bear on the Chinese government to ensure independent and impartial access to political and cultural prisoners. NGOs working in Nepal and North India should be aware that mental health illnesses are prevalent within this community and make efforts to deliver care and ensure appropriate political representation for the displaced individuals. Governments should be aware that human rights violations are unacceptable behavior from an economic partner, and pressure the Chinese government for change. Finally, the Tibetan Government in exile must recognize that mental health illnesses are widespread within the refugee populations and reach out to sympathetic nations for appropriate assistance.
Competing interests
The author(s) declare that they have no competing interest.
Authors' contributions
Concept: EM, SS, TH, SD, JSB, JO
Data searching: EM, SS, TH, SD, JSB
Analysis: EM, SS, TH, SD, JSB, JO
First draft: EM, SS, TH, SD, JSB, JO
Critical revisions: EM, SS, TH, SD, JSB, JO
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors thank Dr. Farah Thong for research assistance.
==== Refs
Servan-Schreiber D Le Lin B Birmaher B Prevalence of posttraumatic stress disorder and major depressive disorder in Tibetan refugee children J Am Acad Child Adolesc Psychiatry 1998 37 874 879 9695450 10.1097/00004583-199808000-00018
UNHCR. Basic Facts. http://www.unhcr.ch/cgi-bin/texis/vtx/basics Accessed May 11, 2005 2005
Fazel M Wheeler J Danesh J Prevalence of serious mental disorder in 7000 refugees resettled in western countries: a systematic review Lancet 2005 365 1309 1314 15823380 10.1016/S0140-6736(05)61027-6
Dolma S Singh S Lohfeld L Orbinski J Mills E The dangerous passage of Tibetan refugees fleeing to Nepal Presented at McMaster University Research Day 2005 Hamilton
United Nations. Convention Against Torture and Other Cruel I, or Degrading Treatment or Punishment. Ga Res 39/46 Annex, 39 UN GAOR, Supp. (no 51) 1984, U.N. Doc. E/CN. 4/1984/72, Annex (1984), art 2 para 1
Mollica RF Harvard Trauma Questionnaire Manual 1987 Boston , Harvard Press
Mollica RF Wyshak G de Marneffe D Khuon F Lavelle J Indochinese versions of the Hopkins Symptom Checklist-25: a screening instrument for the psychiatric care of refugees Am J Psychiatry 1987 144 497 500 3565621
Crescenzi A Ketzer E Van Ommeren M Phuntsok K Komproe I de Jong JT Effect of political imprisonment and trauma history on recent Tibetan refugees in India J Trauma Stress 2002 15 369 375 12392224 10.1023/A:1020129107279
Holtz TH Refugee trauma versus torture trauma: a retrospective controlled cohort study of Tibetan refugees J Nerv Ment Dis 1998 186 24 34 9457144 10.1097/00005053-199801000-00005
Terheggen MA Stroebe MS Kleber RJ Western conceptualizations and Eastern experience: a cross-cultural study of traumatic stress reactions among Tibetan refugees in India J Trauma Stress 2001 14 391 403 11469164 10.1023/A:1011177204593
PHR Striking hard: Torture in Tibet 1997 Boston , http://www.phrusa.org/research/torture/tortib2.html
de Jong JT Komproe IH Van Ommeren M Common mental disorders in postconflict settings Lancet 2003 361 2128 2130 12826440 10.1016/S0140-6736(03)13692-6
Van Ommeren M Sharma B de Jong J Culture, trauma, and psychotrauma programmes Lancet 1997 350 595 9284805 10.1016/S0140-6736(05)63188-1
Bracken PJ Giller JE Summerfield D Psychological responses to war and atrocity: the limitations of current concepts Soc Sci Med 1995 40 1073 1082 7597460 10.1016/0277-9536(94)00181-R
Veijola J Jokelainen J Laksy K Kantojarvi L Kokkonen P Jarvelin MR Joukamaa M The Hopkins Symptom Checklist-25 in screening DSM-III-R axis-I disorders Nord J Psychiatry 2003 57 119 123 12745774 10.1080/08039480310000941
Mollica RF McInnes K Pham T Smith Fawzi MC Murphy E Lin L The dose-effect relationships between torture and psychiatric symptoms in Vietnamese ex-political detainees and a comparison group J Nerv Ment Dis 1998 186 543 553 9741560 10.1097/00005053-199809000-00005
Mollica RF Poole C Son L Murray CC Tor S Effects of war trauma on Cambodian refugee adolescents' functional health and mental health status J Am Acad Child Adolesc Psychiatry 1997 36 1098 1106 9256589 10.1097/00004583-199708000-00017
Shrestha NM Sharma B Van Ommeren M Regmi S Makaju R Komproe I Shrestha GB de Jong JT Impact of torture on refugees displaced within the developing world: symptomatology among Bhutanese refugees in Nepal Jama 1998 280 443 448 9701080 10.1001/jama.280.5.443
Mollica RF McInnes K Sarajlic N Lavelle J Sarajlic I Massagli MP Disability associated with psychiatric comorbidity and health status in Bosnian refugees living in Croatia Jama 1999 282 433 439 10442658 10.1001/jama.282.5.433
Smith Fawzi MC Murphy E Pham T Lin L Poole C Mollica RF The validity of screening for post-traumatic stress disorder and major depression among Vietnamese former political prisoners Acta Psychiatr Scand 1997 95 87 93 9065671
Mouanoutoua VL Brown LG Hopkins Symptom Checklist-25, Hmong version: a screening instrument for psychological distress J Pers Assess 1995 64 376 383 7722862
Porter M Haslam N Predisplacement and postdisplacement factors associated with mental health of refugees and internally displaced persons: a meta-analysis Jama 2005 294 602 612 16077055 10.1001/jama.294.5.602
Mercer SW Ager A Ruwanpura E Psychosocial distress of Tibetans in exile: integrating western interventions with traditional beliefs and practice Soc Sci Med 2005 60 179 189 15482877 10.1016/j.socscimed.2004.04.025
CRIMINAL LAW OF THE PEOPLE'S REPUBLIC OF CHINA. (Adopted at the Second Session of the Fifth National People's Congress on July 1, 1979, promulgated by Order No. 5 of the Chairman of the Standing Committee of the National People's Congress on July 6, 1979, and effective as of January 1, 1980) http://wwwnovexcncom/criminal_lawhtml
Convention against Torture and Other Cruel, Inhuman or Degrading Treatment or Punishment, G.A. res. 39/46, [annex, 39 U.N. GAOR Supp. (No. 51) at 197, U.N. Doc. A/39/51 (1984)], entered into force June 26, 1987.
http://www1umnedu/humanrts/instree/h2catochtm
Convention on the Rights of the Child http://wwwunhchrch/html/menu3/b/k2crchtm 1989
|
16280079
|
PMC1308816
|
CC BY
|
2021-01-04 16:29:56
|
no
|
BMC Int Health Hum Rights. 2005 Nov 9; 5:7
|
utf-8
|
BMC Int Health Hum Rights
| 2,005 |
10.1186/1472-698X-5-7
|
oa_comm
|
==== Front
BMC NephrolBMC Nephrology1471-2369BioMed Central London 1471-2369-6-131630305510.1186/1471-2369-6-13Research ArticleSmoking, oxidative stress and inflammation: Impact on resting energy expenditure in diabetic nephropathy Agarwal Rajiv [email protected] Department of Medicine, Indiana University and RLR VA Medical Center, Indianapolis, Indiana, USA2005 22 11 2005 6 13 13 1 6 2005 22 11 2005 Copyright © 2005 Agarwal; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Inflammation is associated with increased resting energy expenditure (REE) in patients with chronic kidney disease. Oxidative stress, on the other hand, appears not to increase REE. Smoking is a common mechanism for generating oxidative stress and inflammation. Whether smokers have increased REE and if so, whether it is accounted for by the pro-oxidant and inflammatory state is not known.
Methods
A case control study of 11 smokers and 24 non-smokers with overt diabetic nephropathy was performed to evaluate the chronic effect of smoking on REE. REE (indirect calorimetry), glomerular filtration rate (iothalamate clearance), markers of oxidative stress (urinary and plasma malondialdehyde (MDA), and protein carbonyls) and inflammation (C-reactive protein, tumor necrosis factor-alpha, interleukin-6) were measured on two occasions four months apart.
Results
Biomarkers of inflammation (C-reactive protein) and oxidative stress (urinary and plasma MDA) were increased in smokers. REE was increased in smokers, 24.3 kcal/kg/day compared to 21 kcal/kg/day (p = 0.009) in non-smokers. After adjusting for age, GFR, MDA, C-reactive protein, and hemoglobin A1C the difference in REE between the two groups persisted (adjusted difference 3.51 kcal/kg/d, 95% confidence interval 0.59 – 6.45, p = 0.020).
Conclusion
Patients with overt diabetic nephropathy who smoke have a higher REE, oxidative and inflammatory state. Elevated REE is not attributable to heightened oxidative stress and inflammatory state. Smoking is an independent risk factor for elevated REE in patients with diabetic nephropathy and provides an additional mechanism by which it may lead to poor outcomes.
==== Body
Background
Diabetic nephropathy is the commonest cause of end-stage renal disease (ESRD). Malnutrition in patients with ESRD is associated with increased morbidity and mortality. Nutritional decline begins long before patients with progressive CKD become dialysis dependent [1,2] which in part is related to the spontaneous reduction in dietary protein and caloric intake with reduction in glomerular filtration rate (GFR) [3-8].
Recent studies point out that chronic subclinical inflammation in patients with CKD is associated with an increase in REE [9]. Other studies, done in patients without CKD, show that oxidative stress is not associated with increase in REE [10]. Smoking is associated with both an increase in oxidative stress and subclinical inflammation, thus serves as a model to study the effect of oxidative-inflammatory state on REE [11]. Smoking may therefore compound the oxidative stress and the inflammatory state in patients with CKD [2,12-14]. Whether smoking-induced oxidative-inflammatory state is sufficient to account for elevated REE is not known.
Although smoking-induced increase REE is thought to be acute and transient – in part due to nicotine-induced sympathoadrenal activation – more recent studies point to a chronic elevation in REE [15]. In epidemiological studies, increased REE is seen in young women who smoke compared to women who do not, even after controlling for differences in body size and overnight abstinence from smoking [16]. Smoking cessation often leads to weight gain [17]. Thus, smoking may accelerate malnutrition via accelerating attrition in renal function in addition to elevating REE.
In this study we asked the question whether smokers with diabetic nephropathy have an increased REE compared to non-smokers with equally severe diabetic nephropathy. If so, is the increased REE mediated by the pro-oxidant and inflammatory state.
Methods
Subjects
Thirty-five clinically stable Veterans with diabetic nephropathy were recruited from the Renal Clinic of the Roudebush VA Medical Center, Indianapolis, IN after review and approval of protocol by the Institutional Review Board of Indiana University and VA Research and Development Committee. Signed written informed consent was obtained from each participant.
Patients with type 2 diabetes requiring treatment with oral hypoglycemic drugs or insulin, were required to have a urine protein/creatinine ratio of >1.0 g/g on a single voided specimen and a creatinine clearance of >20 mL/min by Cockcroft-Gault formula. All patients were participating in a randomized clinical trial examining the influence of pioglitazone compared to glipizide on reduction of proteinuria. Exclusion criteria included the presence of liver disease, NYHA Class III or IV heart failure, unstable angina, myocardial infarction or stroke in the previous 3 months, NSAID use, or body mass index of ≥40 kg/m2.
Patients underwent a standardized history and physical examination that included assessment of medications and control of diabetes mellitus by hemoglobin A1C. Current smoking status was ascertained by a yes or no response. The following measurements were made in 32 of 35 subjects on two occasions four months apart; in the remaining three only one measurement was made:
Glomerular filtration rate
GFR was measured by iothalamate clearance using a continuous subcutaneous infusion at a rate of 125 μL/hour that was started after a bolus intravenous injection 24 hour prior to actual measurements [18]. The following day, six samples of plasma were collected at times 0, 1.0, 1.5, 2.0, 2.5 and 3.0 hours while the subjects were water loaded. Plasma iothalamate concentrations were analyzed using a previously reported high-performance liquid chromatography technique [19]. The ratio of infusion rate to steady-state plasma concentration yielded the iothalamate clearance. The average of these six collections was expressed as the GFR.
Resting energy expenditure
Seated measurements were made in the morning after an overnight fast. Subjects were asked not to drink or eat anything and refrain from smoking after midnight. They were asked only take half the usual dose of insulin if they were on insulin. Two indirect calorimetric measurements were made 4 months apart in 32 patients and one measurement in the remaining 3 patients.
REE was calculated using open circuit indirect calorimetry in a computerized metabolic system; the gas concentrations of O2 and CO2 were measured in real time using as gas mass-spectrophotometer (Marquette Electronics, Inc, Milwaukee, WI) connected to a computer. The flow rate of the inhaled and exhaled air was simultaneously measured using a flow meter, interfaced to an analog to digital conversion board with data sampled at 20 Hz and stored to a memory file. The data was analyzed using First Breath Software (First Breath Inc., Ontario, Canada) to calculate oxygen consumption and carbon dioxide production in mL/min. The modified Weir's equation (REE (Kcal/day) = [VO2 mL/min (3.94) + VCO2 mL/min (1.11)] × 1.44) [20] was used to calculate the REE. Data were collected for 30 minutes and for the purposes of analysis the first 5 minutes of the collection were discarded since they did not represent steady state.
Lean body mass
To relate REE to differences in body composition between smokers and non-smokers, lean body mass was measured by anthropometric and body impedance spectroscopy.
Anthropometric method
Skinfold thickness measurements were taken in triplicate for biceps, triceps, subscapular and supra-iliac areas by a single observer using Lange skin-fold calipers (Beta Technology Inc, Cambridge, MD, USA) and lean body mass (FFM SFT) was calculated by Durnin and Womersley formula [21].
Body impedance spectroscopy
Body composition was measured using bioelectric impedance analysis (BIA). A BIA-106 Spectrum II analyzer (RJL Systems, Inc., Clinton Twp. MI) was used and data was collected using Weight Manager Version 2.0 software (RJL Systems, Inc., Clinton Twp. MI). The technique of BIA involved introducing an imperceptible 800 μA alternating current at 50 kHz through signal introduction cables placed with pre-gelled electrodes on the ulnar head and medial malleolus. Sensing electrodes placed at the ipsilateral 3rd metacarpal and 2nd metatarsal, determined the electrical resistance and reactance. Lean body mass was calculated from estimates of total body potassium.
Protein nitrogen appearance
Two 24-hour urine samples were collected for the measurement of creatinine, urea, sodium, potassium, chloride and protein. The average of the two 24-hour excretion rates was used for analyses. Dietary protein intake was estimated from the urea nitrogen appearance rate (UNA), calculated as follows [22]: Protein intake (g/day) = 6.25[urine urea nitrogen (UUN) (g/day) + 0.031 × body weight(kg)]
Oxidative stress markers
Malondialdehyde (MDA)
Malondialdehyde (MDA), a lipid hydroperoxide, is formed by β-scission of peroxidized polyunsaturated fatty acids and was measured by high performance liquid chromatography following derivatization with thiobarbituric acid (TBA) as reported previously [23]. Urinary MDA was analyzed in three spontaneously voided specimens over a period of 3 hours. Each specimen was analyzed for MDA and creatinine. The MDA/creatinine ratio was averaged and used further for statistical analysis. Blood was collected in tubes containing EDTA at first and last visits. Plasma was separated by centrifugation and stored at -84°C until analysis
Carbonyl measurements
Oxidation of plasma and urine proteins were measured by analysis of Western blots with slight modifications of a previous report [24]. Total protein was determined using bicinchoninic acid (BCA-1 Protein Assay Kit, Sigma, St. Louis). Urine protein was measured by the dye binding method using a complex of pyrogallol red and molybdenum acid (QuanTtest Red Total Protein Assay System, Quantimetrix, Redondo Beach, CA). Plasma was diluted 1:25 (v:v) with phosphate-buffered saline (PBS), one aliquot of the diluted sample was derivatized and another prepared as an underivatized control using the OxyBlot protein oxidation detection kit (Intergen, Purchase, NY, USA) Urine samples were derivatized with dinitro phenyl hydrazine (DNP) or control reagent similarly except that samples were not diluted prior. Derivatized and underivatized plasma or urine samples were loaded on electrophoresis gels in volumes calculated to give 5 μg protein per sample and electrophoresed according to the method of Laemmli on 4 to 20%gradient SDS-PAGEgels (Bio-Rad, Hercules, CA, USA) for 60 minutes at 200 V. Following electroblotting to 0.2 μ nitrocellulose for 60 volt hours, the membrane was blocked with subsequent immunoblotting using OxyBlot Kit methods and reagents. Bands were visualized with chemiluminescence and captured on film and analyzed densitometrically using dedicated software using a GelLogic 100 System (Kodak, Rochester, NY).
Samples for individual patients before and after therapy (glipizide or pioglitazone), including derivatized and underivatized control, were analyzed on a single Western blot to ensure that responses to therapy were compared under the same analytical conditions. The blot was stained for protein with amido black and the ratio of the density of the carbonyl band vs the corresponding protein band was used to calculate the percent carbonylation of total protein and comparisons before and after therapy made. The carbonyl band corresponding to the molecular weight of albumin was also compared to the corresponding amido black band to generate a ratio to assess the proportion of albumin oxidized.
Inflammation markers
Total leukocyte count (WBC count)
Blood was collected in EDTA containing vacutainers (Beckton-Dickinson) and total leukocyte count calculated by the Coulter method (Coulter STK-S, Coulter Electronics, Inc., Hialeah, FL) by our clinical laboratory.
C-reactive Protein (CRP)
CRP was measured by Cobas Integra 400 autoanalyzer using a particle enhanced turubidimetric assay (Cobas Integra C-Reactive Protein Latex, Roche Diagnostics, Indianapolis, IN). The intra-assay coefficient of variation was 1.8% and the inter-assay coefficient of variation was 2.9% at a mean level of 0.62 mg/dL CRP.
Interleukin 6 (IL-6)
Interleukin 6 (IL-6) was assayed in plasma using a sandwich ELISA (Quantikine® kit for Human IL-6 Immunoassay; R&D Systems, Minneapolis, MN). A standard curve was generated using a linear curve-fit. The correlation coefficient for standards was greater than 0.99 and the lowest detectable limit 0.039 pg/ml in undiluted plasma. The intra-assay coefficient of variation was 7.8% and the inter-assay coefficient of variation was 7.2%.
Tumor Necrosis Factor-α (TNF-α)
Tumor Necrosis Factor-α (TNF-α) was assayed in plasma using a sandwich ELISA (Quantikine® kit for Human TNF-α Immunoassay; R&D Systems, Minneapolis, MN). A standard curve was generated using a linear curve-fit. The correlation coefficient for standards was greater than 0.99 and the lowest detectable limit 0.12 pg/ml in undiluted plasma. The intra-assay coefficient of variation was 5.9% and the inter-assay coefficient of variation was 12.6%.
Urinary Monocyte Chemotactic Protein-1 (MCP-1)
MCP1 was assayed in urine using a sandwich ELISA (Quantikine® kit for Human MCP1 Immunoassay; R&D Systems, Minneapolis, MN). Corrections were made for concentration and values expressed as pg MCP1 per mg creatinine. A standard curve was generated using a four parameter logistic curve-fit. The correlation coefficient for standards was greater than 0.99 and the lowest detectable limit 0.7 pg/ml in 1:2 diluted urine. The intra-assay coefficient of variation was 2.5 ± 3.0% and the inter-assay coefficient of variation was 5.6 ± 4.2%.
Statistical analysis
Differences in baseline characteristics were compared by a Chi-squared or unpaired t-test. To determine the relationship between smoking status and REE a mixed model ANOVA was used to account for correlated observations within subject. The dependent variable was REE/kg body weight; smoking status was used as the fixed effect variable and the subject number as the random effect variable. Clinical, oxidative stress and inflammation markers were used as covariates as noted. The adjusted least square means and 95% confidence intervals obtained form this ANOVA model are reported. SPSS for Windows (version 13, SPSS, Chicago, IL) was used for all analyses. Two sided p-value of less than 0.05 was taken as significant.
Results
The clinical and laboratory characteristics of the study population are presented in Table 1. In keeping with a diagnosis of diabetic nephropathy, the population was obese, required multiple drugs for treatment of hypertension and had a high prevalence of vascular disease. Smokers and non-smokers were well matched, except in that smokers were younger and had higher baseline glomerular filtration rate.
Body composition, as measured by body impedance spectroscopy and skin-fold thickness assessment, was no different between smokers and non-smokers (Table 2).
Table 3 shows markers of oxidative stress, inflammation, and other metabolic parameters. Plasma and urinary MDA were significantly increased in smokers reflecting the heightened state of oxidative stress. However, protein oxidation markers were similar in the two groups. C-reactive protein was also significantly increased in smokers, reflecting the inflammatory state. Plasma interleukin-6 was marginally higher in smokers, however tumor necrosis factor-α did not reach statistical significance. Urine MCP-1 excretion was nearly equal in smokers and non-smokers. Thyroid stimulating hormone reflecting the overall metabolic state was also no different between the smokers and non-smokers. GFR was significantly higher in smokers at 46.7 mL/min/1.73 m2 compared to non-smokers at 29.1 mL/min/1.73 m2.
Body weight is an important determinant of REE in smokers and non-smokers (Figure 1). For any given weight, smokers had a higher REE compared to non-smokers. Unadjusted increase in REE was 3.26 kcal/kg/d in smokers compared to non-smokers (p = 0.009). When adjusted for baseline differences in age, GFR, the oxidative stress markers MDA in urine and plasma, and C-reactive protein did not diminish the elevated REE attributable to smoking (Table 4).
There was no independent relationship between MDA/creatinine ratio and REE (partial correlation coefficient = 0.163 (p = 0.19)) or between CRP and REE (partial correlation coefficient = 0.177, (p = 0.16)) after accounting for correlated measurements. None of the other markers of inflammation or oxidative stress achieved statistical significance.
Discussion
The major finding of this study is that patients with diabetic nephropathy who smoke have an elevated REE after accounting for body weight – the strongest single predictor of REE [25-27]. Smokers had a heightened state of oxidative stress and inflammation. However, neither biomarkers of oxidative stress, nor inflammation had an independent relationship with REE. Furthermore, the association of smoking and elevated REE was not removed when markers of oxidative stress and inflammation were accounted for, strengthening the possibility that oxidative stress and inflammation markers by themselves are insufficient to account for elevated REE in smokers with diabetic nephropathy.
That oxidative stress by itself has little role in elevating REE has been demonstrated in the elderly without kidney disease [10] Our study supports these observations and extends this to patients with CKD.
The relationships between inflammation, creatinine clearance, and REE have been analyzed by Avesani et al in a cross-sectional study of in 91 non-dialyzed patients with CKD [9]. Consideration of kidney function is important for evaluating REE because the oxygen consumption of the kidneys is high – estimated to be between 5.6% and 20% of the total metabolic rate [28,29] – and failing kidneys may have reduced oxygen consumption contributing to reduced REE [29,30]. REE may also decrease as an adaptation to the decreased energy intake or accumulation of uremic toxins that depresses metabolic activity. However, studies are not conclusive regarding this observation. Whereas some studies have shown decreased [31,32] REE others have shown no change [33] or even increased [34] REE in CKD. Avesani et al did not find any relationship between quartiles of creatinine clearance and REE but they were amongst the first to propose inflammation as a cause of elevated REE in CKD [35]. Those in the highest tertile of CRP – and accordingly most inflamed – had an unadjusted REE that was higher compared to the lower tertiles of CRP. Consideration of smoking status as a cause of raised REE has not been reported in any of the above studies. It is possible that the association of increased CRP with REE may simply be due to overrepresentation by smokers in the highest tertile of CRP. Our study suggests consideration of smoking as a covariate in analyses of REE in patients with CKD.
Given that GFR can have a variable influence on REE, as noted above, it is important to account for the effect of GFR in studies dealing with REE in patients with CKD. Smokers in this study had a statistically increased GFR compared to non-smokers which is similar to what has been reported for a large population based study from France [36]. However, adjustment for GFR did not remove the effect of smoking to elevate REE. Other adjustments, such as that for age, CRP, urine and plasma MDA also did not remove the association of increased REE with smoking. Thus, other unmeasured factors may account for the elevated REE in smokers.
The effect of cigarette smoking on REE are thought to be acute and transient and related to sympathetic activation [15]. It should be noted that this study did not examine the acute effect of smoking, rather the chronic effect of smoking. The large effect of smoking on REE was somewhat unexpected. In one epidemiologic study, women who smoked had a statistically significant increase in REE [16]. The magnitude of the effect of smoking was 68 kcal/d which was approximately one-third of the effect size reported in the present study [16]. In another study, black women who smoked were found to have a greater baseline REE compared to white women [37]. Specifically, amongst black women, heavy smoking was associated with 218 kcal/d higher REE and moderate smoking with 177 kcal/d higher REE compared to non-smokers. Although studies on the effect of smoking in diabetic nephropathy or those with CKD have not been reported, the results of this study are in line with those reported by others in populations without CKD [16,37] and extend these observations to patients with CKD. Exactly how chronic smoking leads to increased REE is not clear from present investigations [38].
There are some limitations of this study. First, the sample size was relatively small and limited to men. Nevertheless, repeated observations in the same patient improved the precision of the measurements and demonstrated a highly significant difference between smokers and non-smokers. Second, diabetes control was less than optimal in some patients that can elevate resting energy expenditure by itself. However, adjustment for diabetes control did not mitigate the significance of the results. Third, we did not quantify the amount and duration of smoking that may have an independent effect on REE [37]. Finally, these data may not be applicable to women.
In summary, this study demonstrates higher REE in smokers with diabetic nephropathy after adjusting for body weight. These findings generate the hypothesis that smokers with diabetic nephropathy would be most prone to protein-energy malnutrition at the onset of ESRD and add to other health-risks of smoking in this population[31,33,34,39]. Aside from increasing REE, smoking is known to suppress appetite, which given the anorexic state of uremia, can produce malnutrition by two ways – lowering intake and increasing energy expenditure. Smoking should be considered as a cause of increased REE in patients with CKD when analyzing the effect of other variables such as inflammation and GFR. Longitudinal studies are required to better define the complex relationship between REE, falling renal function, oxidative stress and inflammation in patients with CKD.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
The author designed and secured funding for the study, trained the study personnel in the collection of the data, recruited the patients and finally analyzed the data and wrote the manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The grant support of Takeda Pharma, North America and assistance of Roudebush VA Clinical Chemistry and Microbiology Lab, Indianapolis for CRP and TSH assay is gratefully acknowledged. The author thanks Nadine G. Sachs for the nursing assistance, Shawn D. Chase and Nicole L. Knipe for technical assistance, Raji Bala, Clinical Chemistry supervisor for facilitating the clinical assays and all the patients who gratefully agreed to participate in the research study.
Figures and Tables
Figure 1 Resting energy expenditure is plotted against body weight. A linear relationship is seen for smokers and non-smokers. Inset shows that the mean unadjusted REE in smokers is higher by 3.26 kcal/kg/day (p = 0.009)
Table 1 Clinical and Laboratory Characteristics of Patients with Type 2 Diabetic Nephropathy
Parameter Non-smokers Mean ± SD or N (%) Smokers Mean ± SD or N (%)
N 24 11
Age (y) 69.1 ± 7.0 61.5 ± 9.3 *
Males/Females 24/0 11/0
Race White/Black 19/5 9/2
Weight (kg) 97.8 ± 16.8 102.5 ± 23.4
BMI (kg/m2) 32.6 ± 4.9 32.6 ± 6.6
Serum albumin (g/dL) 3.8 ± 0.33 3.63 ± 0.30
Serum Creatinine (mg/dL) 2.75 ± 1.21 1.80 ± 0.74 *
BUN (mg/dL) 46.5 ± 25.6 38.3 ± 21.8
Hemoglobin (g/dL) 13.3 ± 1.6 13.5 ± 2.4
24 hour urine protein (g/d) 3393 ± 2522 4423 ± 4385
Iothalamate clearance (GFR) (mL/min/1.73 m2) 28.9 ± 13.8 47.2 ± 34.8 *
Duration of diabetes (yr) 15.7 ± 9.1 13.3 ± 9.0
Insulin use 15 (63%) 6 (55%)
Hemoglobin A1C (%) 7.57 ± 2.02 8.98 ± 2.93
Waist/Hip Ratio 1.00 ± 0.07 1.04 ± 0.07
Antihypertensive medications 4.3 ± 1.6 3.9 ± 1.9
ACE inhibitor or Angiotensin receptor blocker use 21 (88%) 8 (73%)
Total Cholesterol (mg/dL) 184 ± 51 223 ± 87
LDL Cholesterol (mg/dL) 100 ± 44 116 ± 24
HDL Cholesterol (mg/dL) 42 ± 11.1 46 ± 11.4
Vascular Disease (coronary, cerebrovascular or peripheral vascular disease) 17 (71%) 7 (64%)
* p < 0.05
Table 2 Body Composition Analysis by Body Impedance Spectroscopy and Skin Fold Thickness
Parameter Non-Smokers N = 24 Smokers N = 11
Body Weight (kg) 99.2 (91.5 – 106.9) 101 (89.6 – 112.4)
Phase angle 5.70 (4.46 – 6.93) 5.12 (3.30 – 6.94)
Vector length 227 (210 – 244) 202 (177 – 227)
Fat free mass (kg) 74.4 (68.2 – 80.6) 81.5 (72.4 – 90.6)
Total body water (L) 55.3 (50.9 – 59.8) 60.6 (54.1 – 67.1)
Fat mass (kg) 25.4 (20.1 – 30.7) 21.3 (13.5 – 29.0)
Total body potassium (mEq) 3981 (3650 – 4313) 4130 (3641 – 4618)
Body cell mass (kg) 33.2 (30.4 – 35.9) 34.4 (30.3 – 38.5)
Extracellular mass (kg) 41.3 (36.4 – 46.1) 46.9 (39.8 – 54.1)
Intracellular water (L) 30.3 (27.7 – 32.8) 31.4 (27.7 – 35.1)
Extracellular water (L) 25.1 (21.6 – 28.5) 29.0 (23.9 – 34.1)
Fat free mass from Skin fold thickness (kg) 53.4 (50.2 – 56.6) 56.4 (51.6 – 61.2)
Least square mean values with 95% confidence intervals are calculated from the mixed model ANOVA model as described in methods. None of the comparisons were significantly different between the groups.
Table 3 Comparison of REE, Inflammation and Oxidative Stress Markers in Smokers and Non-smokers
Parameter Non-smokers Smokers Difference (Smokers – non-smokers) p
REE/kg body wt. 21.0 (19.7 – 22.3) 24.3 (22.3 – 26.2) 3.26 (0.87 – 5.6) 0.009
Respiratory Quotient 0.85 (0.83 – 0.87) 0.81 (0.78 – 0.85) NS
Oxidative Stress Makers
Plasma malondialdehyde (μM/L) 0.94 (0.79 – 1.08) 1.20 (0.98 – 1.41) 0.26 (0.0 – 0.52) 0.05
Urine malondialdehyde/urine creatinine ratio (μM/g creatinine) 3.36 (2.44 – 4.28) 6.33 (4.95 – 7.72) 2.97 (1.31 – 4.64) 0.001
Plasma total carbonyl (DU) 0.818 (0.728 – 0.908) 0.791 (0.658 – 0.924) NS
Urine total carbonyl (DU) 0.454 (0.336 – 0.571) 0.444 (0.270 – 0.617) NS
Plasma albumin carbonyl (DU) 1.425 (1.278 – 1.572) 1.441 (1.225 – 1.658) NS
Urine albumin carbonyl (DU) 0.598 (0.438 – 0.758) 0.591 (0.356 – 0.827) NS
Inflammation Markers
White blood cell count (number/μL) 7897 (6885 – 8908) 9139 (7647 – 10,631) NS
C-reactive protein (mg/dL) 0.80 (0.37 – 1.23) 1.77 (1.13 – 2.40) 0.97 (0.21 – 1.74) 0.014
Interleukin-6 (pg/mL) 2.13 (1.48 – 2.78) 3.28 (2.31 – 4.25) 1.15 (-0.02 – 2.32) 0.053
Tumor necrosis factor-α (pg/mL) 4.05 (3.05 – 5.05) 3.89 (2.40 – 5.38) NS
MCP-1/Creatinine (pg/mg) 398 (288 – 508) 399 (228 – 569) NS
Other variables
HbA1C (%) 7.46 (6.64 – 8.27) 8.73 (7.52 – 9.94) NS
Plasma glucose (mg/dL) 132 (110 – 153) 168 (136 – 200) NS
Protein nitrogen appearance rate (g/kg/d) 0.76 (0.68 – 0.83) 0.81 (0.71 – 0.92) NS
Iothalamate clearance (mL/min/1.73 m2) 29.1 (19.6 – 38.6) 46.7 (32.7 – 60.6) 17.6 (0.7 – 34.6) 0.042
24 hour urine protein (mg/d) 3384 (2180 – 4588) 4189 (2410 – 5967) NS
Thyroid stimulating hormone 2.32 (1.87 – 2.78) 2.04 (1.37 – 2.71) NS
Values reported are least square mean values with 95% confidence intervals are calculated from the mixed model ANOVA model as described in methods. DU stands for densitometric units, REE for resting energy expenditure, MCP-1 for monocyte chemotactic protein-1.
Table 4 Increase in resting energy expenditure in unadjusted and adjusted models associated with smoking
Adjusted variable Resting Energy Expenditure (kcal/kg/d) 95% confidence interval of the estimate P value
None (Unadjusted model) 3.26 0.87 – 5.6 0.009
Age 3.70 1.05 – 6.35 0.008
Age + GFR 3.67 0.94 – 6.40 0.01
Age + GFR + Urine MDA/Cr 3.95 1.06 – 6.84 0.009
Age + GFR + Urine MDA/Cr + CRP 3.71 0.74 – 6.69 0.016
Age + GFR + Urine MDA/Cr + CRP + Plasma MDA 3.48 0.57 – 6.39 0.021
Age + GFR + Urine MDA/Cr + CRP + Plasma MDA + HbA1C 3.51 0.59 – 6.45 0.020
Adjustment for covariates are evaluated at the means which were age = 66.7 years, GFR = 35.6 mL/min/1.73 m2, Urine MDA/Cr = 4.3 μmol/g, Plasma MDA 1.03 μmol/L, CRP = 1.00 mg/dL, HbA1C = 7.78%.
==== Refs
Group PDS Adequacy of dialysis and nutrition in continuous peritoneal dialysis: association with clinical outcomes. Canada-USA (CANUSA) J Am Soc Nephrol 1996 7 198 207 8785388
Stenvinkel P Heimburger O Paultre F Diczfalusy U Wang T Berglund L Jogestrand T Strong association between malnutrition, inflammation, and atherosclerosis in chronic renal failure Kidney Int 1999 55 1899 1911 10231453 10.1046/j.1523-1755.1999.00422.x
Kopple JD Greene T Chumlea WC Hollinger D Maroni BJ Merrill D Scherch LK Schulman G Wang SR Zimmer GS Relationship between nutritional status and the glomerular filtration rate: results from the MDRD study Kidney Int 2000 57 1688 1703 10760105 10.1046/j.1523-1755.2000.00014.x
Lusvarghi E Fantuzzi AL Medici G Barbi L D'Amelio A Natural history of nutrition in chronic renal failure Nephrol Dial Transplant 1996 11 Suppl 9 75 84 9050039
Caravaca F Arrobas M Pizarro JL Sanchez-Casado E Uraemic symptoms, nutritional status and renal function in pre-dialysis end-stage renal failure patients Nephrol Dial Transplant 2001 16 776 782 11274273 10.1093/ndt/16.4.776
Pollock CA Ibels LS Zhu FY Warnant M Caterson RJ Waugh DA Mahony JF Protein intake in renal disease J Am Soc Nephrol 1997 8 777 783 9176847
Heimburger O Qureshi AR Blaner WS Berglund L Stenvinkel P Hand-grip muscle strength, lean body mass, and plasma proteins as markers of nutritional status in patients with chronic renal failure close to start of dialysis therapy Am J Kidney Dis 2000 36 1213 1225 11096047
Ikizler TA Greene JH Wingard RL Parker RA Hakim RM Spontaneous dietary protein intake during progression of chronic renal failure J Am Soc Nephrol 1995 6 1386 1391 8589313
Avesani CM Draibe SA Kamimura MA Colugnati FA Cuppari L Resting energy expenditure of chronic kidney disease patients: influence of renal function and subclinical inflammation Am J Kidney Dis 2004 44 1008 1016 15558521
Bell C Jones PP Seals DR Oxidative stress does not modulate metabolic rate or skeletal muscle sympathetic activity with primary aging in adult humans J Clin Endocrinol Metab 2003 88 4950 4954 14557479 10.1210/jc.2003-030454
Cai H Harrison DG Endothelial dysfunction in cardiovascular diseases: the role of oxidant stress Circ Res 2000 87 840 844 11073878
Raj DS Lim G Levi M Qualls C Jain SK Advanced glycation end products and oxidative stress are increased in chronic allograft nephropathy Am J Kidney Dis 2004 43 154 160 14712439 10.1053/j.ajkd.2003.09.021
Himmelfarb J Stenvinkel P Ikizler TA Hakim RM The elephant in uremia: oxidant stress as a unifying concept of cardiovascular disease in uremia Kidney Int 2002 62 1524 1538 12371953 10.1046/j.1523-1755.2002.00600.x
Agarwal R Chronic kidney disease is associated with oxidative stress independent of hypertension Clin Nephrol 2004 61 377 383 15224800
Perkins KA Metabolic effects of cigarette smoking J Appl Physiol 1992 72 401 409 1559911
Kimm SY Glynn NW Aston CE Poehlman ET Daniels SR Effects of race, cigarette smoking, and use of contraceptive medications on resting energy expenditure in young women Am J Epidemiol 2001 154 718 724 11590084 10.1093/aje/154.8.718
Smoking cessation and severity of weight gain N Engl J Med 1991 325 517 518 1898489
Agarwal R Ambulatory GFR measurement with cold iothalamate in adults with chronic kidney disease Am J Kidney Dis 2003 41 752 759 12666061
Agarwal R Vasavada N Chase SD Liquid chromatography for iothalamate in biological samples J Chromatogr B Analyt Technol Biomed Life Sci 2003 785 345 352 12554148
Weir JB New methods for calculating metabolic rate with special reference to protein metabolism J Physiol 1949 109 1 9 15394301
Durnin JV Womersley J Body fat assessed from total body density and its estimation from skinfold thickness: measurements on 481 men and women aged from 16 to 72 years Br J Nutr 1974 32 77 97 4843734 10.1079/BJN19740060
Maroni BJ Steinman TI Mitch WE A method for estimating nitrogen intake of patients with chronic renal failure Kidney Int 1985 27 58 65 3981873
Agarwal R Chase SD Rapid, fluorimetric-liquid chromatographic determination of malondialdehyde in biological samples J Chromatogr B Analyt Technol Biomed Life Sci 2002 775 121 126 12101069
Agarwal R Proinflammatory effects of oxidative stress in chronic kidney disease: role of additional angiotensin II blockade Am J Physiol Renal Physiol 2003 284 F863 F869 12505865
Nelson KM Weinsier RL Long CL Schutz Y Prediction of resting energy expenditure from fat-free mass and fat mass Am J Clin Nutr 1992 56 848 856 1415003
Leibel RL Rosenbaum M Hirsch J Changes in energy expenditure resulting from altered body weight N Engl J Med 1995 332 621 628 7632212 10.1056/NEJM199503093321001
James WP Davies HL Bailes J Dauncey MJ Elevated metabolic rates in obesity Lancet 1978 1 1122 1125 77416 10.1016/S0140-6736(78)90300-8
Anonymous Comorbid conditions and correlations with mortality risk among 3,399 incident hemodialysis patients Am J Kidney Dis 1992 20 32 38 1442773
Goldwasser P Mittman N Antignani A Burrell D Michel MA Collier J Avram MM Predictors of mortality in hemodialysis patients J Am Soc Nephrol 1993 3 1613 1622 8507818
Panesar A Agarwal R Resting energy expenditure in chronic kidney disease: relationship with glomerular filtration rate Clin Nephrol 2003 59 360 366 12779098
Bucciante G Senesi G Piva M Donato D Velluti F Milito F Spisani C Bernardi A Resting metabolic rate by indirect calorimetry in uremic patients Contrib Nephrol 1990 81 214 219 2093500
Avesani CM Draibe SA Kamimura MA Dalboni MA Colugnati FA Cuppari L Decreased resting energy expenditure in non-dialysed chronic kidney disease patients Nephrol Dial Transplant 2004 19 3091 3097 15507474 10.1093/ndt/gfh547
Schneeweiss B Graninger W Stockenhuber F Druml W Ferenci P Eichinger S Grimm G Laggner AN Lenz K Energy metabolism in acute and chronic renal failure Am J Clin Nutr 1990 52 596 601 2403054
Kuhlmann U Schwickardi M Trebst R Lange H Resting metabolic rate in chronic renal failure J Ren Nutr 2001 11 202 206 11680000 10.1053/jren.2001.26981
Marcussen N Schumann J Campbell P Kjellstrand C Cytodiagnostic urinalysis is very useful in the differential diagnosis of acute renal failure and can predict the severity Ren Fail 1995 17 721 729 8771245
Halimi JM Giraudeau B Vol S Caces E Nivet H Lebranchu Y Tichet J Effects of current smoking and smoking discontinuation on renal function and proteinuria in the general population [In Process Citation] Kidney Int 2000 58 1285 1292 10972692 10.1046/j.1523-1755.2000.00284.x
Clemens LH Klesges RC Slawson DL Bush AJ Cigarette smoking is associated with energy balance in premenopausal African-American adult women differently than in similarly aged white women Int J Obes Relat Metab Disord 2003 27 1219 1226 14513070 10.1038/sj.ijo.0802400
Warwick PM Baines J Energy expenditure in free-living smokers and nonsmokers: comparison between factorial, intake-balance, and doubly labeled water measures Am J Clin Nutr 1996 63 15 21 8604664
Avesani CM Cuppari L Silva AC Sigulem DM Cendoroglo M Sesso R Draibe SA Resting energy expenditure in pre-dialysis diabetic patients Nephrol Dial Transplant 2001 16 556 565 11239031 10.1093/ndt/16.3.556
|
16303055
|
PMC1308817
|
CC BY
|
2021-01-04 16:03:43
|
no
|
BMC Nephrol. 2005 Nov 22; 6:13
|
utf-8
|
BMC Nephrol
| 2,005 |
10.1186/1471-2369-6-13
|
oa_comm
|
==== Front
BMC NeurolBMC Neurology1471-2377BioMed Central London 1471-2377-5-211628009010.1186/1471-2377-5-21Research ArticleRegional variation in hospitalization for stroke among Asians/Pacific Islanders in the United States: a nationwide retrospective cohort study Nguyen-Huynh Mai N [email protected] S Claiborne [email protected] Department of Neurology, University of California San Francisco, 505 Parnassus Avenue, San Francisco, California 94143, USA2 Department of Epidemiology, University of California San Francisco, 505 Parnassus Avenue, San Francisco, California 94143, USA2005 10 11 2005 5 21 21 3 3 2005 10 11 2005 Copyright © 2005 Nguyen-Huynh and Johnston; licensee BioMed Central Ltd.2005Nguyen-Huynh and Johnston; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
In Asia, stroke incidence varies dramatically from country to country. Little is known about stroke incidence in Asians/Pacific Islanders in the US, where regional heterogeneity in Asian/Pacific Islander sub-populations is great. We sought to characterize both the national and regional incidences of first and recurrent hospitalized acute ischemic stroke, subarachnoid hemorrhage, and intracerebral hemorrhage in Asians/Pacific Islanders compared to non-Hispanic whites.
Methods
We used the National Inpatient Sample of the 1997 Healthcare Cost and Utilization Project. It is a 20% stratified sample of hospitalizations to nonfederal hospitals in the US. National and regional projections were made using sampling weights specific for patients and hospitals. We identified stroke subtypes using previously validated ICD-9 codes. Age-adjusted incidence rates were calculated using the direct method with the US population in 2000 as the standard.
Results
There were 169,386 stroke hospitalizations in the database. Nationally, compared to whites, Asians/Pacific Islanders were more likely to have subarachnoid hemorrhage (incidence rate ratio {RR} female: 1.53, 95% CI 1.41–1.65; male RR: 1.13, 95% CI 1.00–1.27) and intracerebral hemorrhage (female RR 1.29, 95% CI 1.22–1.36; male RR: 1.58, 95% CI 1.50–1.67). However, when examined by geographic regions, Asians/Pacific Islanders had higher incidence rates of subarachnoid hemorrhage and intracerebral hemorrhage predominantly in the West, and lower rates of stroke elsewhere.
Conclusion
Stroke incidence varies 3-fold among Asians/Pacific Islanders residing in different US regions. Geographic variation is less dramatic in whites. Whether genetic or cultural differences are responsible for dramatic heterogeneity among Asian/Pacific Islander populations is unclear and deserves further study.
==== Body
Background
Stroke is a disease of major importance in Asia. Reported incidence rates vary dramatically but are generally higher than those of the US [1-3]. For example, reported incidence rates for overall stroke are 39% greater in Japan [4], 23% greater in Taiwan [5], and 81% greater in Northern China[6] compared to the US. It is the second leading cause of death in China, Korea and Taiwan; third in Japan and Singapore; sixth in the Philippines; and tenth in Thailand [3,7-9].
The incidence of stroke among Asians/Pacific Islanders (APIs) in the US is largely unstudied. The only major epidemiologic study on stroke in a sub-population of APIs was the Honolulu Heart Program [10]. A recent comparison of the 20-year incidence of hospitalized stroke between Japanese-American men in the Honolulu Heart Program and white men in the Framingham Study found a 40% excess of thromboembolic stroke in whites [11]. The incidence of hemorrhagic stroke was similar in the two cohorts. Another study reported a higher mortality rate from hemorrhagic stroke in APIs compared to whites in the US, and that deaths tended to occur at younger ages for some stroke types among APIs [12]. Some small studies have reported differences in the distribution of stroke subtypes between APIs and whites in the US [13,14]. APIs have constituted only a small proportion of prior population-based cohorts in the US.
Although race-ethnic disparities within the US has been a recent interest [15], the vast majority of studies have focused on African Americans and Hispanics [16,17], and few studies have described stroke in APIs in the US. API constitutes approximately 4% of the total US population and represents the fastest growing racial-ethnic group in the country. The Bureau of Census estimates that the number of Asians aged 25 years or older will increase 5-fold from 1995 to 2050 and that API will reach 10% of the US population by the year 2050 [18]. In 1998, the Department of Health and Human Services launched the Initiative to Eliminate Racial and Ethnic Disparities in Health with stroke identified as one of the six targeted conditions, and Healthy People 2010 prevention agenda[19] has the same objective.
Examining race-ethnic distributions of stroke incidence and mortality is required to define the burden of disease and to help guide the development of directed prevention programs, and may also be useful in identifying new genetic or environmental risk factors. We sought to characterize the incidence of hospitalized acute ischemic stroke, subarachnoid hemorrhage (SAH), and intracerebral hemorrhage (ICH) in APIs compared to non-Hispanic whites, using data representative of US hospitalizations in 1997. To assess whether stroke rates might differ in API sub-populations in the US, we also characterized regional variation.
Methods
We used the Nationwide Inpatient Sample of the 1997 Healthcare Cost Utilization Project (HCUP), release 6, produced by the Agency for Healthcare Research and Quality [20]. The Nationwide Inpatient Sample (NIS) is a sample of US community hospitals, including nonfederal, short-term, general and specialty hospitals. It excludes all long-term hospitals, psychiatric hospitals, and alcoholism/chemical dependency treatment facilities. The NIS is based on a 20% stratified probability sample of hospitals with sampling probabilities proportional to the total number of US community hospitals in each stratum. The five stratification variables are geographic region (Northeast, Midwest, South, and West), ownership (government nonfederal, private not-for-profit, private investor-owned), location (urban vs. rural), teaching status (teaching vs. non-teaching), and size based on number of beds (small, medium and large). The database includes information typically available from discharge abstracts for 7,148,420 inpatient stays in 1,012 hospitals across 22 geographically dispersed states (AZ, CA, CO, CT, FL, GA, HI, IL, IA, KS, MD, MA, MO, NJ, NY, OR, PA, SC, TN, UT, WA, WI). Our study was approved by the Committee on Human Research at the University of California San Francisco.
Incidence of hospitalized strokes
HCUP does not distinguish between first and recurrent stroke admissions; therefore, we report the incidence of first and recurrent stroke leading to hospitalization. We identified all patients with a primary diagnosis of stroke (International Classification of Diseases, Ninth Revision, [ICD-9] codes 430–434, 436). Stroke subtypes were identified from previously identified codes [21-23]: acute ischemic stroke (ICD-9 433.01, 433.11, 433.21, 433.31, 433.81, 433.91, 434.01, 434.11, 434.91, 436), SAH (ICD-9 430), and ICH (ICD-9 431). For standardization of stroke incidence, age groups were defined as 0–19 years, 20–29, 30–39, 40–49, 50–59, 60–64, 65–74, 75–84, and 85+. Gender was classified as male and female. Race-ethnicity was coded as non-Hispanic white, Black, Hispanic, API, Native American, and other, based on self-report.
Unadjusted incidence rates were calculated using the number of cases in the database, the appropriate discharge sampling weights specific for each age group, stroke type, and race-ethnicity, and with the 1997 non-Hispanic API or white populations based on Census Bureau data [24] as the denominators. Age-adjusted and gender-specific incidence rates were calculated by the direct method using the entire US population in 2000 as the standard [25]. The number of exposed individuals for each age category was calculated by dividing the 1997 US population in each age category by the discharge sampling weight for the corresponding age category. Subsequently, an event rate for each age category was calculated by dividing the number of cases identified in the database by the number of exposed individuals. Age adjusted rate equals to the sum of age-specific rates multiplied by the age-specific US population proportions. Standard errors for all age categories were computed, as well as an overall standard error for the age-adjusted incidence rate. A 95% confidence interval (CI) was equal to the adjusted rate ± 1.96 times the standard error.
Estimates were obtained for national as well as regional incidences. The four geographical regions were Northeast, Midwest, South, and West as defined by the US Census Bureau in 2000 [26]. Point estimates and 95% CI for incidence rate ratios were calculated using non-Hispanic white as the reference group. Variances for adjusted incidence rates were calculated and regional heterogeneity was tested using ANOVA. Pair wise t-test with Bonferroni correction for multiple comparisons was used to determine statistical differences between pairs of regional incidence rates among APIs and non-Hispanic whites.
Case fatality
Case fatality, or in-hospital death due to all causes, was calculated for each stroke subtype in non-Hispanic whites, and APIs. The Wilcoxon rank-sum test was used to evaluate race-ethnic differences in age, and the chi-squared test was used to test for differences in stroke subtype. Robust logistic regression [27] was used to calculate the independent association between API race and stroke case fatality. The robust method broadens CI to account for clustering of covariates and outcomes by hospital. In multivariable analyses, we adjusted for patient characteristics such as age, sex, co-morbidity score, length of stay and median income for patient's zip code. Co-morbidity scores were developed using a database version of the Charlson comorbidity index and represent a summary of major secondary diagnoses weighted by severity [28]. We also adjusted for hospital characteristics including geographic region, location (urban vs. rural), teaching status (teaching vs. non-teaching), ownership (government non-federal, private for profit, private non-profit), and size (small, medium, large). To determine whether there was heterogeneity among API populations within the US, we tested for interactions between ethnicity and region by using the likelihood-ratio test [29]. The Stata statistical package was used for all analyses (version 7.0; Stata Corporation, College Station, TX).
Results
There were 169,386 hospitalizations for stroke in the 1997 NIS database. Of these, 102,268 (60.4%) were for acute ischemic stroke, 5,172 (3.1%) for SAH and 14,907 (8.8%) for ICH. An additional 47,039 (27.7%) were for non-specific stroke diagnoses, with carotid stenosis or occlusion (ICD-9 433.1) being the most frequent. Using the average discharge weight for each stroke subtype, we calculated the US projections to be 510,113 cases of acute ischemic stroke, 24,893 cases of SAH, and 73,208 cases of ICH. Most patients were admitted through the emergency department (57.9%). The distribution of race-ethnicity was 80.9% white, 1.6% API, 11.4% African American, 4.7% Hispanic, 0.2% Native American, and 1.3% other. The overall average age was 71.7 ± 13.4 years, and 52.7% were female.
Compared to whites, APIs hospitalized with strokes were younger (Table 1). On average, APIs lived in areas with higher median incomes, were less likely to be discharged to home, had longer hospital stays, and incurred higher total charges. The baseline characteristics for each race-ethnic group were also examined by stroke subtype. The results remained similar to those for all strokes combined in Table 1a, except that APIs with SAH had shorter length of stay than whites. The mean ± SD age for whites were 74.9 ± 12 years for those hospitalized with ischemic stroke, 59.3 ± 16.5 years for SAH, and 72.2 ± 14 years for ICH. APIs were younger across all three stroke subtypes: 71 ± 12.2 years for ischemic stroke, 57.9 ± 18.7 for SAH, and 65 ± 15.3 for ICH. Admissions for ICH made up a larger percentage of total strokes among APIs (17.4%) compared to non-Hispanic whites (8.1%), which could account for some differences in mortality outcomes and age at presentation since patients with ICH tend to be younger [30]. In addition, some of the patient characteristics for API varied significantly across the four regions (Table 2). These include type of hospital, median income for patient's zip code, length of stay, expected payer, and discharge status.
Table 1 Characteristics of all hospitalized stroke patients by race-ethnicity.
Asian/Pacific Islanders
Mean ± SD or no.(%) (n = 2,159) non-Hispanic Whites
Mean ± SD or no. (%) (n = 109,632)
Age, years 68.5 ± 14.3 73.3 ± 12.3
Female 47.3 52.1
Region
Northeast 224 (10.4) 26,471 (24.2)
Midwest 52 (2.4) 21,022 (19.2)
South 186 (8.6) 43,080 (39.3)
West 1697 (78.6) 19,059 (17.4)
Urban hospital 2019 (93.5) 93,538 (85.5)
Teaching hospital 821 (38.0) 31,039 (28.4)
Type of hospital
Government, non-fed 432 (20.0) 12,829 (11.7)
Private, not-for-profit 1423 (65.9) 38,748 (72.0)
Private, for-profit 304 (14.1) 17,794 (16.3)
Median income for patient's zip code
$0 – $25,000 254 (12.3) 31,077 (30.3)
$25,001 – $35,000 649 (31.5) 38,748 (37.7)
$35,001 + 1155 (56.1) 32,876 (32.0)
Admit from Emergency Dept 1450 (67.4) 58,970 (56.2)
Length of stay, days 8.2 ± 30.4 5.8 ± 19.4
Expected payer
Medicare 1017 (47.1) 81,956 (74.8)
Medicaid 451 (20.9) 2740 (2.5)
Private 542 (25.1) 21,435 (19.6)
Other 149 (6.9) 3402 (3.1)
Total charges $24,026 ± $34,928 $14,275 ± $21,039
Discharge status
Routine 789 (36.6) 52,698 (48.1)
Home health care 202 (9.4) 8443 (7.7)
Other facility/AMA* 1130 (43) 39,376 (36.1)
Died 238 (11.0) 8891 (8.1)
Average age at death, years 68.3 ± 14.5 75.5 ± 13.0
* AMA = against medical advice
Table 2 Characteristics of all hospitalized Asian patients with stroke by region.
Asian/Pacific Islanders
Mean ± SD or no.(%) (n = 2,159)
Northeast
(n = 224) Midwest
(n = 52) South
(n = 186) West
(n = 1697)
Age, years 64.3 ± 15.97 64.7 ± 15.7 66 ± 14.8 69.2 ± 14
Female 44.2 40.4 43.6 48.4
Urban hospital 222 (99.1) 46 (88.5) 184 (98.9) 1567 (92.3)
Teaching hospital 160 (71.4) 19 (36.5) 35 (18.8) 607 (35.8)
Type of hospital
Government, non-federal 64 (28.6) 8 (15.4) 24 (12.9) 336 (19.8)
Private, not-for-profit 155 (69.2) 44 (84.6) 127 (68.3) 1097 (64.6)
Private, for-profit 5 (2.2) 0 35 (18.8) 264 (15.6)
Median income for patient's zip code
$0 – $25,000 21 (9.6) 27 (52.9) 26 (14.8) 180 (11.2)
$25,001 – $35,000 96 (44.0) 11 (21.6) 36 (20.5) 506 (31.4)
$35,001 + 101 (46.3) 13 (25.5) 114 (64.8) 927 (57.5)
Admit from Emergency Dept 168 (75.3) 34 (72.3) 140 (75.7) 1108 (65.4)
Length of stay, days 16.2 ± 87.5 6.1 ± 7.2 7.8 ± 7.9 8.1 ± 17.3
Expected payer
Medicare 75 (33.5) 14 (26.9) 78 (41.9) 850 (50.1)
Medicaid 50 (22.3) 17 (32.7) 35 (18.8) 349 (20.6)
Private 65 (29.0) 14 (26.9) 49 (26.3) 414 (24.4)
Other 34 (15.2) 7 (13.5) 24 (12.9) 84 (4.9)
Total charges $25,570 ± $31,001 $14,172 ± $27,282 $15,151 ± $17,994 $25,161 ± $36,920
Discharge status
Routine 95 (42.8) 24 (46.2) 90 (48.4) 580 (34.2)
Home health 17 (7.7) 7 (13.5) 15 (8.1) 163 (9.6)
Other facility/AMA 82 (36.9) 17 (32.6) 64 (34.4) 765 (45.1)
Died 28 (12.6) 4 (7.7) 17 (9.1) 189 (11.1)
Average age at death, years 69.3 ± 14.4 59.3 ± 23.4 68.6 ± 12.8 68.4 ± 14.5
* AMA = against medical advice
We calculated the national age-adjusted and gender-specific incidence rates for hospitalized acute ischemic stroke, SAH, and ICH. Compared to non-Hispanic whites, APIs were more likely to have SAH and ICH, but rates of acute ischemic stroke were similar (Table 3). There was significant regional variation in incidence rates. API females had higher incidence rate ratios for SAH (1.53, 95% CI 1.41–1.65) and ICH (1.69, 95% CI 1.60–1.79) in the West, but lower rates of stroke everywhere else in the country compared to non-Hispanic whites. Similarly, API males had higher incidence rate ratios for acute ischemic stroke (1.16, 95% CI 1.12–1.21), SAH (1.34, 95% CI 1.12–1.60), and ICH (2.02, 95% CI 1.87–2.18) in the West. In addition, API males also had higher incidence rate ratios for SAH in the Northeast (1.31, 95% CI 1.05–1.63), and ICH in the South (1.21, 95% CI 1.06–1.38) compared to non-Hispanic whites. Regional variation was less dramatic for whites. There was great heterogeneity in incidence rates across regions within each stratum of stroke type (Table 3). In addition, pair-wise t-test calculations revealed that rates for APIs in the West were significantly different from those of all other regions (p < 0.008, after Bonferroni correction).
Table 3 Age-adjusted incidence rates and rate ratios of hospitalized stroke stratified by gender for Asians/Pacific Islanders and non-Hispanic whites by stroke subtype and US region.*
Incidence Rate per 100,000 (95% CI) Incidence Rate Ratio (95% CI)
Asians/Pacific Islanders Non-Hispanic Whites
Female Male Female Male Female Male
Ischemic Stroke
National 111 (102–120) 106 (98–114) 151 (149–153) 131 (129–133) 0.74 (0.72–0.76) 0.81 (0.78–0.83)
Northeast 72 (50–94) 63 (46–80) 187 (183–191) 162 (158–166) 0.39 (0.36–0.41) 0.39 (0.36–0.42)
Midwest 36 (14–58) 35 (18–52) 116 (113–119) 98 (95–101) 0.31 (0.27–0.36) 0.36 (0.31–0.41)
South 79 (55–104) 84 (62–106) 171 (168–170) 158 (155–161) 0.46 (0.43–0.49) 0.53 (0.50–0.57)
West 120 (109–131) 118 (108–128) 128 (125–131) 101 (98–104) 0.94 (0.91–0.98) 1.16 (1.12–1.21)
Heterogeneity p < 0.001 p < 0.001 p < 0.001 p < 0.001
SAH
National 12 (9–15) 6 (4–8) 8 (7.6–8.4) 5 (4.7–5.3) 1.53 (1.41–1.65) 1.13 (1.00–1.27)
Northeast 5 (0–10) 8 (3–13) 9 (7–10) 6 (5–7) 0.56 (0.41–0.74) 1.31 (1.05–1.63)
Midwest 1 (0–2) 2 (0–5) 7 (6–8) 4 (3–5) 0.09 (0.02–0.23) 0.63 (0.34–1.07)
South 2 (0–4) 1 (0–2) 8 (7–9) 5 (4–6) 0.28 (0.18–0.42) 0.12 (0.04–0.26)
West 16 (12–20) 6 (4–8) 8 (7–9) 5 (4–6) 1.92 (1.73–2.14) 1.34 (1.12–1.60)
Heterogeneity p < 0.001 p = 0.008 p = 0.079 p < 0.001
ICH
National 25 (21–29) 30 (26–34) 20 (19–21) 19 (18–20) 1.29 (1.22–1.36) 1.58 (1.50–1.67)
Northeast 21 (11–25) 19 (11–27) 25 (24–26) 24 (23–25) 0.86 (0.75–0.98) 0.77 (0.66–0.88)
Midwest 8 (0–18) 12 (2–22) 16 (15–17) 14 (13–15) 0.53 (0.40–0.70) 0.87 (0.68–1.11)
South 20 (10–30) 25 (13–37) 22 (21–23) 21 (20–22) 0.94 (0.82–1.08) 1.21 (1.06–1.38)
West 27 (22–32) 35 (30–40) 19 (18–20) 17 (16–18) 1.42 (1.31–1.54) 2.02 (1.87–2.18)
Heterogeneity p = 0.039 p = 0.001 p < 0.001 p < 0.001
* SAH = subarachnoid hemorrhage, ICH = intracerebral hemorrhage. Incidence rate ratio is for Asians/Pacific Islanders vs. non-Hispanic whites. Variances for adjusted rates were calculated and regional heterogeneity was tested using ANOVA.
Case fatality
Among all stroke hospitalizations, in-hospital death occurred in 14,024 (8.3%). The mean ± SD age for those who died in-hospital was 73.4 ± 14.3 years for all races. APIs (68.3 ± 14.5 years) died at a younger age than non-Hispanic whites (75.5 ± 13 years, p < 0.001). These race-ethnic differences in age at death remained when each stroke subtype was examined separately (data not shown). Case fatality for acute ischemic stroke was 7.3% in white and 6.4% in API (p < 0.001). Case fatality for SAH was 27.5% in white, 29.8% in API (p = 0.081). Case fatality for ICH was 31.9% in white, 26.7% in API (p = 0.002).
We assessed the odds of in-hospital death by stroke subtypes for APIs as compared to whites. In univariate analyses, APIs (OR = 0.78, 95% CI 0.61–0.99, p = 0.05) admitted with ICH were less likely to die in-hospital. However, the association was no longer significant after adjustment for patient and hospital characteristics (age, sex, co-morbidity score, length of stay, and median income for patient's zip code, geographic region, location, teaching status, ownership, and bed size). In addition, there was no interaction between race-ethnicity and region in case fatality (p = 0.50).
Discussion
Few studies have examined stroke in APIs in the US and no population-based study on stroke in API is available. Thus, the incidence of stroke in API in the US is not known. Using a 1997 national database to overcome problems with small sample size, we found that the overall US incidence rates for SAH and ICH were higher for APIs than for non-Hispanic whites. Our age-adjusted and gender-specific incidence rates for total stroke and for each stroke subtype for non-Hispanic whites are similar to those reported elsewhere[31,32]. The age and gender-adjusted incidence of first-ever strokes among whites in Rochester was 179 per 100,000[31]. A recent population-based study of the greater Cincinnati metropolitan area reported the age and gender-adjusted incidence rates for non-Hispanic whites to be 212 per 100,000 for first and recurrent strokes, 137 per 100,000 for first ischemic stroke, 6 per 100,000 for first SAH, and 18 per 100,000 for first ICH[32].
We found a significant interaction between race-ethnicity and geographic region. The stroke incidence rates varied over 3-fold in APIs among different US regions. Geographic variation was less dramatic in non-Hispanic whites. It is not likely that the heterogeneity in rates was due to small numbers given that the incidence rates were statistically heterogeneous among the regions for all stroke types. This dramatic regional variation might reflect the different distributions of subgroups of APIs (such as Chinese, Japanese, Filipinos, etc.) in the US, as suggested by the significant differences in patient characteristics across regions (Table 2). Striking regional differences in stroke incidence between northern and southern China [3], and between countries in Asia[7] have been reported. These reports and our current findings suggest that combining these distinct subpopulations into one race-ethnic category may be inappropriate.
Our finding of a higher incidence of ICH in APIs is consistent with prior reports from Asia of high rates of hemorrhagic strokes in Asians [3,5,33]. Studies from Asia also reveal a high prevalence of hypertension in these populations [33]. It is not known how much uncontrolled hypertension contributes to the differential burden of hemorrhage in APIs in the US. Other differences in diet (such as a higher salt intake), prevalence of smoking, or genetic factors might also contribute to this difference in ICH incidences.
On average, APIs had longer hospital stays for ischemic stroke and ICH, had higher total hospital charges, and were more likely to be discharged to another facility rather than home. These findings were not explained by differences in stroke subtypes. One possible explanation of these findings is that APIs are hospitalized with more severe strokes, as has been suggested by one prior study showing higher case fatality rates for APIs with ICH [12]. However, we found no significant difference in all-cause in-hospital mortality between API and non-Hispanic white after adjusting for important potential confounders. Whether cultural differences in preferences for end-of-life care could mask an impact of greater stroke severity on case fatality cannot be answered in our study.
There are several limitations to our study. We only included in our analyses patients who were hospitalized for stroke. Thus, non-hospitalized strokes, which account for approximately 10% of all strokes in the US [34], were not included. Furthermore, incidence-rate calculations were based on all stroke hospitalizations including recurrent cases. The accuracy of our incidence estimates depends on the accuracy of ICD-9 coding in the database. Although we used only ICD-9 codes that were previously identified as more accurate in order to enhance the precision of our estimates, reliability of coding is still imperfect [21-23]. In addition, by using only ICD-9 codes in the primary position, we recognize that we increase specificity but decrease the sensitivity of case ascertainment [35]. Another limitation is the lack of reliable coding for comorbidities. Consequently, we were unable to fully adjust for severity of disease or to evaluate risk factors. We also did not have data on do-not-resuscitate status of the patients, which could influence the in-hospital mortality rate. Furthermore, API was not broken down into distinct ethnic subgroups in the database. This is a problem encountered by all similar studies. Despite the limitations, the study reveals several important and interesting findings that deserve further exploration.
Conclusion
In summary, analyses from a large nationwide database revealed that APIs appear to have a higher risk than non-Hispanic whites for ICH and SAH. However, differences between APIs and whites vary dramatically between regions, with the highest rates among APIs in the West. Which sub-populations of APIs are at greater risk, and the factors that predict this elevated risk remain to be determined. APIs come from nearly 50 distinct countries and ethnic groups, each with its own set of cultures and traditions, and potentially very different genetic and/or environmental risk factors for stroke. Our findings suggest the need for further studies to examine the differences in stroke incidence and risk factors among the subpopulations of APIs. Furthermore, understanding of the racial-ethnic differences in stroke should lead to the development of more efficient and effective stroke prevention strategies.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MNH and SCJ both contributed to the conception and design of the study. MNH conducted the analyses of the data and drafted the manuscript. Both authors made substantial contributions to the revision of the manuscript, and approved the final version.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
MNH was supported by AHA grant 0475019N. SCJ was supported by NIH/NINDS grant K02 NS 02254.
==== Refs
Thom TJ Epstein FH Feldman JJ al. Total mortality and mortality from heart disease, cancer, and stroke from 1950 to 1987 in 27 countries. 1992 Washington DC, National Institutes of Health
Gordon T Mortality experience among the Japanese in the United States, Hawaii, and Japan Public Health Rep 1957 72 543 553 13432134
Wong KS Haung YN Gao S Lam WWM Chan YL Cerebrovascular disease among Chinese populations--recent epidemiological and neuroimaging studies. Hong Kong Med J 2001 7 50 57 11406676
Morikawa Y Nakagawa H Naruse Y Nishijo M Miura K Tabata M Hirokawa W Kagamimori S Honda M Yoshita K Hayashi K Trends in Stroke Incidence and Acute Case Fatality in a Japanese Rural Area -- The Oyabe Study Stroke 2000 31 1583 1587 10884458
Hu HH Sheng WY Chu FL Lan CF Chiang BN Incidence of Stroke in Taiwan Stroke 1992 23 1237 1241 1519277
Cheng XM Ziegler DK Lai YH Stroke in china, 1986 through 1990. Stroke 1995 26 1990 1994 7482636
Causes of Death - Southeast Asian Medical Information Center, 1999 SEAMIC Health Statistics
Annual report on births and deaths 1970 to 1994. Registry of Births and Deaths, Singapore
Lee SB Roh JK Yoon BW Hong SB Lee JH Epidemiology of cerebrovascular disease in korea. A collaborative study, 1989-1990 J Korean Med Sci 1993 8 281 289 8198765
Kagan A Harris BR Winkelstein WJ Johnson KG Kato H Syme SL Rhoads GG Gay ML Nichaman MZ Hamilton HB Tillotson J Epidemiologic studies of coronary heart disease and stroke in japanese men living in japan, hawaii and california: Demographic, physical, dietary, and biochemical characteristics J Chronic Dis 1974 27 345 364 4436426 10.1016/0021-9681(74)90014-9
Rodriguez BL D'Agostino R Abbott RD Kagan A Burchfiel CM Yano K Ross GW Silbershatz H Higgins MW Popper J Wolf PA Curb JD Risk of hospitalized stroke in men enrolled in the Honolulu Heart Program and the Framingham study. A comparison of incidence and risk factor effects Stroke 2002 33 230 236 11779915 10.1161/hs0102.101081
Ayala C Greenlund KJ Croft JB Keenan NL Donehoo RS Giles WH Kittner SJ Marks JS Racial/ethnic disparities in mortality by stroke subtype in the United States, 1995-1998 Am J Epidemiol 2001 154 1057 1063 11724723 10.1093/aje/154.11.1057
Caplan LR Gorelick PB Hier DB Race, sex and occlusive cerebrovascular disease: a review Stroke 1986 17 648 655 3526645
Feldmann E Daneault N Kwan E Ho KJ Pessin MS Langenberg P Caplan LR Chinese-white differences in the distribution of occlusive cerebrovascular disease Neurology 1990 40 1541 1545 2215945
Initiative to Eliminate Racial and Ethnic Disparities in Health -- Department of Health and Human Services
Gaines K Burke G Ethnic differences in stroke: black-white differences in the United States population. SECORDS investigators. Southeastern Consortium on Racial Differences in Stroke Neuroepidemiology 1995 14 209 239 7477665
Gillum RF Stroke in blacks Stroke 1988 19 1 9 3276035
US Bureau of the Census. Population projections of the United States by age, sex, race, and Hispanic origin: 1995-2050. In: Current Population Reports Series P25-1130. 2000 Washington, DC, US Bureau of the Census
US Department of Health and Human Services. Healthy People 2010. 2nd ed. With understanding and improving health and objectives for improving health. 2000 Washington, DC, US Government Printing Office
Healthcare Cost and Utilization Project (HCUP), 1988-2000: A Federal-State-Industry Partnership in Health Data. July 2003. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/data/hcup/hcup-pkt.htm
Ellejaer H Holmen J Kruger O Terent A Identification of incident stroke in Norway: hospital discharge data compared with a population-based stroke register Stroke 1999 30 56 60 9880388
Goldstein LB Accuracy of ICD-9-CM coding for the identification of patients with acute ischemic stroke: effect of modifier codes Stroke 1998 29 1602 1604 9707200
Leibson CL Naessens JM Brown RD Whisnant JP Accuracy of hospital discharge abstracts for identifying stroke Stroke 1994 25 2348 2355 7974572
CDC Wonder -- Census Data Request
Szklo M Nieto FJ Epidemiology Beyond the Basics 2000 Gaithersburg, Aspen Publishers, Inc. 494
Census 2000 Geographic Terms and Concepts. U.S. Census Bureau.
Stata Reference Manual, Release 7, Volume 2 H-P 2001 College Station, Texas, Stata Press 227 228
D'Hoore W Sicotte C Tilquin C Risk adjustment in outcome assessment: the Charlson comorbidity index. Methods Inf Med 1993 32 382 387 8295545
Hosmer DW Lemeshow S Applied Logistic Regression 1989 first New York, John Wiley & Sons
Johnston SC Selvin S Gress DR The burden, trends, and demographics of mortality from subarachnoid hemorrhage Neurology 1998 50 1413 1418 9595997
Brown RD Whisnant JP Sicks JD O'Fallon WM Wiebers DO Stroke incidence, prevalence, and survival: secular trends in Rochester, Minnesota, through 1989 Stroke 1996 27 373 380 8610298
Kissela B Schneider A Kleindorfer D Khoury J Miller R Alwell K Woo D Szaflarski J Gebel J Moomaw C Pancioli A Jauch E Shukla R Broderick J Stroke in a biracial population: the excess burden of stroke among blacks Stroke 2004 35 426 431 14757893 10.1161/01.STR.0000110982.74967.39
Lawes CM Rodgers A Bennett DA Parag V Suh I Ueshima H MacMahon S Asia Pacific Cohort Studies Collaboration. Blood pressure and cardiovascular disease in the Asia Pacific region. J of Hypertension 2003 21 707 716
Williams GR Incidence and characteristics of total stroke in the United States BMC Neurology 2001 1 2 11446903 10.1186/1471-2377-1-2
Tirschwell DL Longstreth WTJ Validating administrative data in stroke research Stroke 2001 33 2465 2470 12364739 10.1161/01.STR.0000032240.28636.BD
|
16280090
|
PMC1308818
|
CC BY
|
2021-01-04 16:28:53
|
no
|
BMC Neurol. 2005 Nov 10; 5:21
|
utf-8
|
BMC Neurol
| 2,005 |
10.1186/1471-2377-5-21
|
oa_comm
|
==== Front
BMC NeurosciBMC Neuroscience1471-2202BioMed Central London 1471-2202-6-641628393410.1186/1471-2202-6-64Research ArticleNeurophysiological correlates of mismatch in lexical access Friedrich Claudia K [email protected] University of Konstanz, Department of Linguistics, Universitaetstrasse 10, P.O.Box D25, D-78457, Konstanz, Germany2 University of Hamburg, Department of Biological Psychology and Neuropsychology, Von-Melle-Park 11, D-20146 Hamburg, Germany2005 11 11 2005 6 64 64 5 7 2005 11 11 2005 Copyright © 2005 Friedrich; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
In the present study neurophysiological correlates related to mismatching information in lexical access were investigated with a fragment priming paradigm. Event-related brain potentials were recorded for written words following spoken word onsets that either matched (e.g., kan – Kante [Engl. edge]), partially mismatched (e.g., kan – Konto [Engl. account]), or were unrelated (e.g., kan – Zunge [Engl. tongue]). Previous psycholinguistic research postulated the activation of multiple words in the listeners' mental lexicon which compete for recognition. Accordingly, matching words were assumed to be strongly activated competitors, which inhibit less strongly activated partially mismatching words.
Results
ERPs for matching and unrelated control words differed between 300 and 400 ms. Difference waves (unrelated control words – matching words) replicate a left-hemispheric P350 effect in this time window. Although smaller than for matching words, a P350 effect and behavioural facilitation was also found for partially mismatching words. Minimum norm solutions point to a left hemispheric centro-temporal source of the P350 effect in both conditions. The P350 is interpreted as a neurophysiological index for the activation of matching words in the listeners' mental lexicon. In contrast to the P350 and the behavioural responses, a brain potential ranging between 350 and 500 ms (N400) was found to be equally reduced for matching and partially mismatching words as compared to unrelated control words. This latter effect might be related to strategic mechanisms in the priming situation.
Conclusion
A left-hemispheric neuronal network engaged in lexical access appears to be gradually activated by matching and partially mismatching words. Results suggest that neural processing of matching words does not inhibit processing of partially mismatching words during early stages of lexical identification. Furthermore, the present results indicate that neurophysiological correlates observed in fragment priming reflect different aspects of target processing that are cumulated in behavioural responses. Particularly the left-hemispheric P350 difference potential appears to be closely related to fine-grained activation differences of modality-independent representations in the listeners' mental lexicon. This neurophysiological index might guide future studies aimed at investigating neural aspects of lexical access.
==== Body
Background
Spoken word comprehension requires a listener to identify a single word among thousands of representations stored in her mental lexicon. Behavioral research suggests that the incoming auditory signal activates multiple lexical representations that match the signal [1], but also representations with partial mismatch to the input [2,3]. Pseudowords like gabinet or mabinet, for example, have been shown to activate the word cabinet. However, this activation is decreased as compared to a complete match. Thus, the matching of input and stored representations results in a specific activation pattern. Words that completely match the input are strongly activated. Words that partially mismatch the input are less strongly activated. Competition among activated candidate words has been postulated as a mechanism that reduces the number of activated words [4,5]. Strongly activated words inhibit less strongly activated words.
Both, graded lexical activation and competition, have been previously investigated using word fragment priming. A word fragment, which is commonly the onset of a spoken word, is immediately followed by a visual word or a meaningless letter string (pseudoword). Participants are asked to decide whether they saw a word or not. Faster responses for words that match the fragment as compared to unrelated control words have been found. For example, responses to music are faster when it follows mus than when it follows viba. This facilitation has been interpreted as reflecting the activation that modality-independent representations of matching words receive from the fragments [6,7] (see [8] for an introduction into from priming paradigms).
Different amounts of overlap between fragment and word modulate reaction times in fragment priming. For example, responses are faster when a fragment has the same stress as the target word than when fragment and target word differ in stress [9,10]. A word like music, with stress on the first syllable, is responded to faster when it is preceded by stressed mus than when it is preceded by unstressed mus. This result illustrates graded lexical activation depending on goodness-of-fit between fragment and word. Furthermore, inhibition has been found when fragments were taken from competitor words. Responses are faster when a word is preceded by an unrelated fragment than when the same word is preceded by a partially mismatching fragment for which a better completion exists [11]. For example, abon taken from the Spanish word abonado (Engl. subscriber) inhibits processing of abanico (Engl. fan).
Event-related brain potentials (ERPs) recorded for targets in fragment priming reveal a previously undescribed left-hemispheric brain potential [12]. It is called the P350 and differentiates matching words from unrelated control words. Difference waves resulting from the subtraction of matching words from unrelated control words reveal more positive amplitudes with a maximum peak at 350 ms. Comparable P350 difference waves have been shown for visual fragments preceding visual words. Therefore, it was concluded that the P350 reflects neural processing related to the identification of modality-independent lexical representations. Furthermore, subtle differences in the speech signal such as pitch contour modulate the P350 effect. Pitch contour is an important parameter that marks lexical stress. P350 effects reveal that words with a stress pattern that matches the pitch contour of the input are stronger activated than words that do not match in their stress pattern [13].
Besides the P350 effect, enhanced N400 amplitudes are reliably found for unrelated control words in fragment priming. The N400 is a frequently observed ERP component in different language related tasks (see [14] for a review). It has been argued that the N400 amplitude does not reflect automatic lexical activation in priming tasks [15-17]. It appears more plausible from a vast amount of N400 results that the amplitude of this component is inversely related to the effort needed to integrate an incoming word in a preceding context formed by a sentence or a single priming stimulus. Accordingly, the N400 might rather be related to post-lexical matching and integration processes than to lexical activation in fragment priming [12]. A lexical account to the N400 in word fragment priming is also challenged by the fact that its amplitude is not sensitive to fine-grained activation differences as a function of pitch [13]. Therefore, we interpreted the N400 as a correlate of strategic effects that are assumed to assist lexical decisions in a priming situation [18-20]. Possible mechanisms that speed up yes-responses to matching words might be a rough phonological matching between prime and target or the expectation of a phonological form, which is established by the fragment. A phonological account to the N400 is supported by N400 reduction found for rhyming words [21-23] or word stem priming [24,25].
The present study aimed to explore ERP correlates of the activation of partially mismatching words following fragments for which better completions exist. Monosyllabic word onset fragments were extracted from German word pairs that had identical first syllables except of the vowel (e.g., Kante [Engl. Edge] and Konto [Engl. account]). Three types of fragment-word pairs summarized in Table 1 were tested: In a Match condition, words were preceded by completely matching fragments. In a Mismatch condition, the same words were preceded by fragments extracted from their pair members. In Control conditions, fragments were followed by unrelated control words. Neural correlates of graded lexical activation are shown. However, neither neurophysiological data nor behavioral results reveal inhibition of partially mismatching words during this early phase of lexical identification.
Results and discussion
Reaction times and error rates
Mean reaction times and error rates are shown in Figure 1A. Behavioural measures were subjected to one-way ANOVAs with the three-level factor Relatedness (matching words vs. partially mismatching words vs. unrelated control words). A main effect of Relatedness allowed further step-down analyses of differences between conditions, F(2,46) = 41.84, Greenhouse-Geisser epsilon = 0.90, corrected p < .001. Responses to completely matching words were faster than responses to unrelated control words, t(1,23) = 58.81, p < .001. Similarly, responses to partially mismatching words were faster than responses to unrelated control words, t(1,23) = 6.04, p = .02. Nevertheless, subjects' responses were faster for matching than for partially mismatching words, t(1,23) = 46.01, p < .001. The same pattern of results was observed for error rates for which again a main effect of Relatedness was observed, F(2,46) = 17.10, Greenhouse-Geisser epsilon = 0.76, corrected p < .001. Subjects made less errors for matching words than for unrelated control words, t(1,23) = 21.82, p < .001. Similarly, subjects made less errors for partially mismatching words than for unrelated control words, t(1,23) = 7.89, p < .01. Nevertheless, responses to matching words were more accurate than responses to partially mismatching words, t(1,23) = 14.90, p < .001.
In sum, reactions were faster and more accurate for matching words than for partially mismatching words. This finding goes in line with behavioral results of previous fragment priming experiments [6,7,9-11]. However, the fact that responses for partially mismatching words were not slower than responses for unrelated control words does not replicate previous work. Remember that in a former study subjects responded faster to unrelated control words (e.g., indi – abanico) than to partially mismatching words (e.g., abon – abanico). This has been taken as evidence that better matching completions inhibit partially mismatching words [11]. The present results indicate that inhibition of partially mismatching words is not an obligatory finding in word fragment priming.
Taken together present and previous work, it might be concluded that the lengths of the fragments is a factor that modulates inhibition effects. In contrast to the former study, in which disyllabic fragments were presented as primes [11], monosyllabic fragments were used in the current experiment. Accordingly, the present results suggest that effective competition needs (i) information exceeding the first syllable of a spoken word and/or (ii) time to exert inhibitory effects. With respect to (i): Single syllables as used in the present experiment might fully and partly activate a large number of competitors that do not effectively inhibit each other. In contrast, disyllabic fragments as used in the previous study might fully activate only a few or at least only one word, which effectively inhibits partly activated words. With respect to (ii): Additional processing time provided by disyllabic fragments, which have a longer duration than monosyllabic fragments, might stabilize inhibitory effects. Further research has to explore both possible influences on competition effects.
ERPs
Mean amplitudes for matching words, partially mismatching words and unrelated control words are shown in Figure 1B for eight selected electrode sites, respectively. Waveforms were characterized by an N1-P2 complex followed by negativity between 200 and 400 ms, most prominent over frontal electrode positions. Amplitudes of left hemispheric electrodes were sensitive to the experimental manipulation in the time range between 300 and 400 ms. Difference waves showed characteristic P350 effects (see Figure 1C). Analysis of P350 effects was identical to a former study with syllabic fragments [13]. Starting at approximately 350 ms, an N400 component was observed over bilateral posterior electrode positions. The N400 effect began earlier and was of shorter duration than that observed in previous fragment priming studies [12,13]. It was examined using a time window between 350 and 500 ms.
P350: 300 to 400 ms
A three-way ANOVA with factors Relatedness (matching words vs. partially mismatching words vs. unrelated control words), Hemisphere (left vs. right electrode positions), and Region (anterior vs. posterior electrode positions) was applied to analyze P350 effects. This analysis yielded significant interactions of the factors Relatedness, and Hemisphere, F(2,46) = 16.07, Greenhouse-Geisser epsilon = 0.98, corrected p < .001; and Relatedness, Region and Hemisphere, F(2,46) = 10.78, Greenhouse-Geisser epsilon = 0.75, corrected p = .01. Amplitudes for matching words differed from unrelated control words over both left hemispheric ROIs, both t(1,23) ≥ 5.26, both p ≤ .03. This suggests lexical activation of matching words. Partially mismatching words also showed some degree of lexical activation. Their amplitudes differed from amplitudes of control words over the left anterior ROI, t(1,23) = 5.92, p = .02. However, the fact that amplitudes for matching words differed from partially mismatching words over both left hemispheric ROIs indicates strongest activation for matching words, both t(1,23) ≥ 11.92, both p ≤ .002.
To sum up, ERPs elicited over the left hemisphere were sensitive to the experimental manipulation in the time window of the previously reported P350 deflection [12,13]. Over left-temporal scalp regions, a positive-going effect with enhanced amplitudes for unrelated control words as compared to matching words replicates former P350 results (see electrodes TP7, CP5, P7, P5 in Figure 1B). In contrast, negative-going ERPs with enhanced amplitudes for matching words were observed in the time window of the P350 over left-anterior regions (see electrodes F7, F5, FT7, FC5 in Figure 1B). What remains stable across studies is that subtraction of ERPs for matching or partially mismatching words from ERPs for unrelated control words results in positive-going difference waves with a maximum at 350 ms. Therefore, it appears more appropriate to apply the label 'P350 effect' to these difference waves than to a positive-going deflection in the ERPs.
P350 effects appear to differ with respect to their scalp topography on an anterior-to-posterior dimension over the left hemisphere. In the present study P350 effects were pronounced over temporo-frontal electrode positions. In contrast, only temporal electrodes were involved in P350 effects found in previous studies [12,13]. An obvious cause for different P350 topographies might be the earlier onset of the N400 effect in the present study as compared to the previous studies. The N400 also shows a posterior scalp distribution, but counteracts the P350 effect in polarity of the elicited differences. The earlier beginning of the N400 effect in the present study may have canceled out the posterior part of the P350 effect resulting in the observed temporo-frontal scalp topography.
The present results are consistent with the idea that P350 effects are closely correlated to the activation status of lexical entries in a modality-independent mental lexicon. Matching words, which are strongly activated by the input, elicit large P350 difference waves. Partially mismatching words, which are less strongly activated by the input, elicit reduced P350 difference waves. However, from the present and the previous results it is difficult to distinguish whether the P350 effect reflects a surplus in positivity for unrelated control words or a surplus in negativity for matching words. As already discussed P350 effects were related to different ERP deflections across the present study and previous studies. It has been suggested that reduced positive-going amplitudes for matching words are a correlate of facilitated lexical identification resulting from pre-activation of lexical entries [12,13]. In contrast, enhanced negative ERP amplitudes for matching words over left temporo-frontal electrodes, which were observed in the present study, might suggest that P350 effects directly result from the activation status of lexical entries.
Negative-going left temporo-frontal ERPs in the time window of the P350 effect look very similar to an N330 effect reported earlier [26]. The N330, which also shows a temporo-frontal scalp distribution, was found to be enhanced for words primed by semantic associates as compared to unprimed words. This contrasts the N400 priming effect classically showing reduced amplitudes for words with semantic relation to their preceding primes (see [14]). Both the N330 effect in semantic priming and the P350 effect in word fragment priming can be functionally separated from the N400 effect. Interestingly enough, the N330 was interpreted as a correlate of semantic or lexical activation. This is in line with the present interpretation of the P350 difference wave being related to lexical activation. It remains to be investigated whether the neural activation mechanism that is reflected in the P350 difference wave might also underlie the previously reported N330.
Minimum norm calculations were conducted to elucidate the localization of neural processes underlying P350 difference waves. Results suggest a left centro-temporal origin of neural sources engaged in the processing of both matching and partially mismatching words (see Figure 1D). Crucially, minimum norm solutions in both conditions differ only in strength of dipole activation but not in estimated neural sources. These results clearly point to a unique underlying left-temporal source for processes that are involved in the mapping of speech input onto modality-independent lexical representations. The underlying neuronal network was stronger activated by matching than by partially mismatching words. The minimum norm solutions suggest a relation between the P350 effect in the ERP and an M350 effect found in magnetic brain responses [27]. The M350 has been characterized as an automatic early component, which appears to be related to lexical access in visual word processing [28]. Time ranges of P350 difference waves and M350 are comparable [12]. Furthermore, source analyses for both neurophysiological correlates point to left-temporal neural activation. Both, the P350 and the M350, might index automatic spreading activation across lexical entries. They provide new means to explore aspects of early word processing and the underlying neuronal mechanisms.
N400: 350 to 500 ms
N400 effects were analyzed using the same statistical design as for P350 effects (see previous section). The three-way ANOVA yielded significant interactions of the factors Relatedness and Region, F(2,46) = 31.86, Greenhouse Geisser epsilon = 0.69, corrected p < .001, Relatedness and Hemisphere, F(2,46) = 5.52, Greenhouse Geisser epsilon = 0.88, corrected p < .01, and Relatedness, Region, and Hemisphere F(2,46) = 9.21, Greenhouse Geisser epsilon = 0.68, corrected p < .01. For both posterior ROIS matching words elicited reduced amplitudes of the N400 as compared to unrelated control words, both t(1,23) = 6.18, both p = .02. Similarly, partially mismatching words elicited reduced amplitudes of the N400 over both posterior ROIs, both t(1,23) ≥ 5.89, both p ≤ .03. However, N400 amplitudes for matching and partially mismatching words did not differ significantly, t(1,23) ≤ 1.52, n.s. Thus, in contrast to P350 difference waves and reaction times, N400 amplitude did not differentiate matching words and partially mismatching words.
The present N400 results go in parallel with the earlier finding that the N400 is not sensitive to a mismatch between the pitch of a fragment and the stress pattern of a succeeding word [13]. Taken together, results of both studies indicate an insensitivity of the N400 to subtle differences between fragment and word. They support an interpretation of the N400 as a correlate of neural processes that operate at a post-lexical level, and indicate that behavioral responses in fragment priming are modulated by more than lexical activation. Phonological matching has been postulated as a possible mechanism that is reflected in the N400 in fragment priming [13]. Note that both, matching words and partially mismatching words, are phonologically related to the fragments (e.g., kon-KONTO or kan-KONTO), whereas unrelated control words show no phonological relation to their primes (e.g., kon-SALTO). The fact that both, matching words and partially mismatching words, elicit reduced N400 amplitudes confirms the assumption that processes underling the N400 provide a superficial matching or expectation that speeds up yes-responses to words with some phonological relation to the fragment. A phonological account to the N400 is also suggested by N400 reduction for rhyming words [21-23] or for word stem priming [24,25].
Conclusion
The present results indicate that fragmentary word information modulates different aspects of neural target processing. This is reflected by two separate ERP correlates, namely P350 and N400 effects. Both can be distinguished as separate neurophysiological deflections in accordance to several relevant characteristics (see for example [29]). They differ with respect to latency, scalp topography, polarity of the elicited differences, and sensitivity to the experimental manipulation. Therefore, P350 and N400 effects probably reflect at least two different neuronal processes, which both precede and modulate behavioral responses in fragment priming. P350 effects on the one hand appear to be related to the fine grained mapping of the acoustic input onto lexical representations. N400 effects on the other hand might be related to phonological matching between fragment and word. This argumentation leads to the conclusion that behavioral data, which are preceded by P350 and N400 effects, reflect outcomes of both processes. ERPs, in particular P350 effects, allow investigating lexical activation more directly than behavioral data.
Both, reaction times and P350 difference waves, suggest that partially mismatching lexical representations, which diverge from spoken word fragments only in the vowel of the first syllable, receive slight activation from the speech input. In contrast to earlier findings [11], both measures reveal that partially mismatching representations are not inhibited by better matching completions. Thus, the present data do not support a strong notion of competition between activated lexical candidates at early stages of lexical processing. They reveal that the human speech recognition system does not strictly inhibit alternative candidates from early lexical activation. Factors that modulate competition effects, such as the length of the fragments, have to be focus of future research.
Finally, the neuronal sources underlying the activation of modality-independent lexical representations can be studied with fragment priming in more detail. ERPs in word fragment priming indicate that activation of representations that receive input from both, spoken and written words, appears to be a left-hemispheric function with underlying neuronal sources in the centro-temporal cortex. Although these neuronal sources appear to be separate from written word from representations in the left fusiform gyrus [30], a relation to speech representations in the left superior temporal sulcus [31] can be assumed. In light of the primacy of auditory language comprehension in ontogenetic development, the question emerges whether modality-independent lexical representations are identical to speech representations. The P350 effect seems to be a powerful neurophysiological means to address this and related questions regarding neuronal bases of lexical access in human language comprehension.
Methods
Participants
Twenty-four right-handed volunteers (12 females, 12 males, mean age 22 years) from the Leipzig Max Planck Institute of Human Cognitive and Brain sciences subject pool took part in the experiment. All subjects were right-handed native speakers of German with no reported hearing or neurological problems and normal or corrected-to-normal vision.
Stimuli and procedures
Sixty word pairs served as carrier for prime fragments. Words in a word pair shared the same first syllable with exception of the vowel (e.g. Kan.te and Kon.to, dots indicate the end of the extracted fragment). To control for co-articulation, the phoneme following the fragment was identical for words in a word pair. All carrier words were spoken by a female native speaker of German and recorded using an analog recorder. Items were then digitized at a sampling rate of 44 kHz with 16-bit analog-to-digital conversion on a PC computer using the software Computerized Speech Lab (CSL, ® Kay Elemetric Corp.). Fragments were created from the digitized signals with CoolEdit (® Syntrillum Software Corp.). All fragments were taken from initially stressed words.
Written forms of the carrier words were presented as target words. Matching words were preceded by the prime extracted from their spoken form (e.g., kon – Konto, 120 trials). The same words also served as partially mismatching words. In the latter case they were preceded by fragments extracted from the respective pair member (e.g., kan – Konto, 120 trials). Control words were completely unrelated to the fragments in word onset (e.g.,kon – Salto, 240 trials). Unrelated control words were matched to the related words with respect to stress, number of syllables and number of letters, as well as with respect to word frequency, according to an online data base of German [32]. For example, Salto served as control word for Konto, whereas Zunge served as control word for Kante. In order to equalize the number of repetitions and to balance related and unrelated trials, unrelated control words were presented twice: Once in combination with the fragment from the matched target (kon-Zunge), and once in combination with the fragment taken from the targets' pair member (kan-Zunge).
In half of the trials pseudowords were presented. Pseudowords were created by interchanging the last one or two letters of two related words or two unrelated words. Pseudoword creation followed phonotactic rules of German. Pseudowords were combined with fragments in the same way as words: They matched the fragment (e.g., kon – Konta, 120 trials), mismatched in the vowel (e.g., kan – Konta, 120 trials) or were unrelated to the fragment (e.g., kon – Salte, 240 trials). The first presentation of targets was counterbalanced for the match, mismatch, and unrelated control conditions, as well as for the words and pseudowords.
Participants were comfortably seated in an electrically and acoustically shielded chamber. Visual stimuli were presented on a computer screen in front of the subjects. An experimental trial began with the presentation of a fixation cross at the centre of the screen. Participants were instructed to fixate this cross whenever it appeared. While the fixation cross remained on the screen, a spoken word onset fragment was presented via loudspeakers after 300 ms. Fragments had a mean duration of 321 ms (SD = 61). Loudspeakers were placed at the left and the right side of the screen. The fixation cross was replaced by a visual word or pseudoword in immediate succession to the auditory fragment. Visual stimuli were presented for 200 ms in uppercase white letters on a black background. The task was to judge as quickly and as correctly as possible whether or not the target was a word. Half the subjects made yes-responses via button press with the thumb of their left hand and no-responses via button press with the thumb of their right hand, for the remaining subjects response hands were reversed. Reaction times were measured from stimulus onset with a time-out of 1500 ms. The next trial started after a fixed inter-trial interval of 1000 ms after the behavioral response, or 2500 ms after onset of the visual stimulus if no response was given.
Electrophysiological recording and data analysis
The EEG was recorded continuously (250Hz/22 bit sampling rate; DC amplifier by Twente Medical Systems) from 58 Ag-AgCl electrodes mounted in an elastic cap (Electro Cap International, Inc.) according to the international 10–10 system. All electrodes were referenced against the nose tip. Impedances were kept below 5 kO. Four further electrodes provided biploar recordings of the horizontal and vertical electro-occulogram (EOG). The vertical EOG was recorded from electrodes placed above and below the right eye; the horizontal EOG was recorded from electrodes at the outer canthus of each eye. An electrode placed at the sternum served as ground. Artifacts caused by facial and eye movements were rejected off-line when one of the EOG recordings exceeded a voltage change of 30 μV or more within 200 ms. Furthermore, a visual inspection of the raw EEG was carried out to eliminate drifts. ERPs were computed for the words starting from the beginning of the visual presentation up to 600 ms and with a 200 ms prestimulus baseline. Only correctly answered, artifact-free trials were included in the averaging procedure. For illustrative purposes only, the grand-average ERPs were smoothed off-line using a 10-Hz lowpass filter.
Minimum norm estimations were calculated using the Source Analysis module of BESA (® MEGIS Software GmbH [33]). Grand average difference waves for the match and the mismatch condition were subjected to the Source analysis module as a whole segment. No additional filtering was applied to the source analysis. Depth weighting and spatio-temporal weighting (Dale & Sereno [34]) were used to calculate source solutions. Noise was estimated using the baseline of the match condition.
Responses shorter than 200 ms, and longer than 1500 ms were removed from both behavioral and ERP analyses. Percent of erroneous responses and reaction times calculated from the onset of the visual words to the subjects' responses were subjected to a one-way ANOVA with the three-level factor Relatedness (matching words vs. partially mismatching words vs. unrelated control words). Degrees of freedom for the three-level factor Relatedness were adjusted by using the Greenhouse-Geisser epsilon. Two additional factors served for analysis of ERP effects. To analyze lateral electrodes, the factor Hemisphere (left vs. right electrode positions) was included. To analyze anterior vs. posterior effects the factor Region (anterior vs. posterior electrode sites) was included. This resulted in four ROIs including 11 electrodes each (anterior left: F9, F7, F5, F3, FT9, FT7, FC5, FC3, T7, C5, C3; anterior right: F10, F8, F6, F4, FT10, FT8, FC6, FC4, T8, C6, C4; posterior left: TP9, TP7, CP5, CP3, P9, P7, P5, P3, PO7, PO3, O1; posterior right: TP10, TP8, CP6, CP4, P10, P8, P6, P4, PO8, PO4, O2). In case of significant interactions, t-tests were computed to evaluate differences among conditions. Only main effects of the factor Relatedness and interactions including this factor and leading to significant post-hoc comparisons are reported. Note, that the time windows and the ROIs applied to analyze the present ERP data were identical to that used in a former fragment priming study with syllabic fragments [13].
Acknowledgements
The work was supported by the Leibniz Prize awarded to Aditi Lahiri. I thank Angela D. Friederici, Carsten Eulitz and Kai Alter for critical discussions on the work presented here, and Isabella Paul for comments on a former version of the manuscript. Special thank goes to Maren Schmidt-Kassow for her assistance in preparation of the stimulus material and in data collection.
Figures and Tables
Figure 1 A) The first part of the figure shows reaction times in milliseconds and error rates in percent for words preceded by completely matching fragments (light blue), for the same words preceded by partially mismatching fragments (pink), and for unrelated control words (grey). Note that error bars represent standard deviations. B) Grand mean averages representing ERPs for 16 selected electrode places over the right and left hemisphere are plotted in this part of the figure (matching words in light blue; mismatching words in pink; unrelated control words in grey solid and dotted lines). Electrode positions are illustrated in the map. The time window for which statistical analyses of the P350 effect were conducted is highlighted in grey. C) Difference waves for four selected left-hemispheric electrode places representing the P350 effect are shown in this part of the figure (unrelated control words – matching words in light blue; unrelated control words – mismatching words in pink). Again the time window of statistical analysis of the P350 effect is highlighted in grey. D) This part of the figure shows minimum norm solutions estimating the neural sources underlying the P350 difference waves for matching words (above) and for mismatching words (below). Colors represent strength of dipole activation in percent relative to the highest level of activation in the match condition at the P350 peak (352 ms).
Table 1 This table displays the experimental design of the present study, and gives examples of trials that were realized in the conditions. The same words were presented as related words, once in combination with a matching fragment, one in combination with a partially mismatching fragment. Similarly, unrelated words were presented twice to control for responses biases and repetition effects. Both types of control words are shown in the figures in order to illustrate that there were virtually no differences between both conditions. Statistical analyses included only one group of control words.
Related words
Match Mismatch
kon-KONTO kan-KONTO
kan-KANTE kon-KANTE
Σ = 120 trials Σ = 120 trials
Unrelated control words
kon-SALTO kan-SALTO
kan-ZUNGE kon-ZUNGE
Σ = 120 trials Σ = 120 trials
==== Refs
Zwitserlood P The locus of the effects of sentential-semantic context in spoken-word processing Cognition 1989 32 25 64 2752705 10.1016/0010-0277(89)90013-9
Connine CM Titone D Deelman T Blasko DG Similarity mapping in spoken word recognition: Evidence from phoneme monitoring Journal of Memory and Language 1997 37 463 480 10.1006/jmla.1997.2535
Frauenfelder UH Scholten M Content A Bottom-up inhibition in lexical selection: Phonological mismatch effects in spoken word recognition Language and Cognitive Processes 2001 16 583 607 10.1080/01690960143000146
Vitevitch MS Luce PA When words compete: Levels of processing in spoken word perception Psychological Science 1998 9 325 329 10.1111/1467-9280.00064
McQueen JM Norris D Cutler A Competition in spoken word recognition: Spotting words in other words Journal of Experimental Psychology: Learning, Memory, and Cognition 1994 20 621 638 10.1037/0278-7393.20.3.621
Marslen-Wilson W Altmann GTM Activation, competition and frequency in lexical access Cognitive models of speech processing: Psycholinguistic and computational perspectives 1990 Massachusetts, MIT Press 148 172
Spinelli E Segui J Radeau M Phonological priming in spoken word recognition with bisyllabic targets Language and Cognitive Processes 2001 16 367 392 10.1080/01690960042000111
Zwitserlood P Form priming Language and Cognitive Processes 1996 11 589 596 10.1080/016909696386980
Cooper N Cutler A Wales R Constraints of lexical stress on lexical access in English: Evidence from native and nonnative listeners Language and Speech 2002 45 207 228 12693685
Cutler A Van Donselaar WA Voornaam is not (really) a homophone: Lexical prosody and lexical access in Dutch Language and Speech 2001 44 171 195 11575903
Soto-Faraco S Sebastián-Gallés N Cutler A Segmental and suprasegmental mismatch in lexical access Journal of Memory and Language 2001 45 412 432 10.1006/jmla.2000.2783
Friedrich CK Kotz SA Friederici AD Gunter TC ERPs reflect lexical identification in word fragment priming Journal of Cognitive Neuroscience 2004 16 541 552 15165347 10.1162/089892904323057281
Friedrich CK Kotz SA Friederici AD Alter K Pitch modulates lexical identification in spoken word recognition: ERP and behavioral evidence Cognitive Brain Research 2004 20 300 308 15183401 10.1016/j.cogbrainres.2004.03.007
Kutas M Federmeier KD Electrophysiology reveals semantic memory use in language comprehension Trends Cognitive Science 2000 4 463 470 11115760 10.1016/S1364-6613(00)01560-6
Holcomb PJ Semantic priming and stimulus degradation – implications for the role of the N400 in language processing Psychophysiology 1993 30 47 61 8416062
Chwilla DJ Brown CM Hagoort P The N400 as a function of the level of processing Psychophysiology 1995 32 274 285 7784536
Chwilla DJ Hagoort CM Brown CM The mechanism underlying backward priming in a lexical decision task: Spreading activation versus semantic matching The Quarterly Journal of Experimental Psychology: Section A 1998 51 531 560
Neely JH Keefe DE Ross KL Semantic priming in the lexical decision task: Roles of prospective prime-generated expectancies and retrospective semantic matching Journal of Experimental Psychology: Learning, Memory, and Cognition 1989 15 1003 1019 2530303 10.1037/0278-7393.15.6.1003
Norris D McQueen JM Cutler A Bias effects in facilitatory phonological priming Memory and Cognition 2002 30 399 411 12061760
Radeau M Morais J Dewier A Phonological priming in spoken word recognition: Task effects. Memory and Cognition 1989 17 525 535
Praamstra P Meyer AS Levelt WJM Neurophysiological manifestations of auditory phonological processing: Latency variation of a negative ERP component timelocked to phonological mismatch Journal of Cognitive Neuroscience 1994 6 204 219
Rugg MD Barrett SE Event-related potentials and the interaction between orthographic and phonological information in a rhyme-judgment task Brain and Language 1987 32 336 361 3690257 10.1016/0093-934X(87)90132-5
Dumay N Benraïss A Barriol B Colin C Radeau M Besson M Behavioral and electrophysiological study of phonological priming between bisyllabic spoken words Journal of Cognitive Neuroscience 2001 13 121 143 11224913 10.1162/089892901564117
Münte TF Say T Clahsen H Schiltz K Kutas M Decomposition of morphologically complex words in English: Evidence from event-related brain potentials Cognitive Brain Research 1999 7 241 253 9838144 10.1016/S0926-6410(98)00028-7
Rodriguez-Fornells A Münte TF Clahsen H Morphological priming in Spanish verb forms: An ERP repetition priming study Journal of Cognitive Neuroscience 2002 14 443 454 11970803 10.1162/089892902317361958
Nobre AC McCarthy G Language-related ERPs: Scalp distributions and modulation by word type and semantic priming Journal of Cognitive Neuroscience 1994 6 233 255
Pylkkänen L Stringfellow A Marantz A Neuromagnetic evidence for the timing of lexical activation: An MEG component sensitive to phonotactic probability but not to neighborhood densitiy Brain and Language 2002 81 666 678 12081430 10.1006/brln.2001.2555
Pylkkänen L Marantz A Tracking the time course of word recognition with MEG Trends in Cognitive Sciences 2003 7 187 189
Kutas M Van Petten CK Gernsbacher MA Psycholinguistics electrified: Event-related brain potential investigations Handbook of Psycholinguistics 1994 San Diego: Academic Press 161 163
McCandliss DB Cohen L Dehaene S The visual word form area: expertise for reading in the fusiform gyrus Trends in Cognitive Sciences 2003 7 293 299 12860187 10.1016/S1364-6613(03)00134-7
Scott SK Johnsrude IS The neuroanatomical and functional organization of speech perception Trends in Neurosciences 2003 26 100 107 12536133 10.1016/S0166-2236(02)00037-1
Wortschatz Deutsch
BESA Brain Electrical Source Analysis
Dale AM Sereno M Improved Localization of Cortical Activity by combining EEG and MEG with cortical Surface Reconstruction: A linear approach Journal of Cognitive Neuroscience 1993 5 162 176
|
16283934
|
PMC1308819
|
CC BY
|
2021-01-04 16:03:49
|
no
|
BMC Neurosci. 2005 Nov 11; 6:64
|
utf-8
|
BMC Neurosci
| 2,005 |
10.1186/1471-2202-6-64
|
oa_comm
|
==== Front
BMC Palliat CareBMC Palliative Care1472-684XBioMed Central London 1472-684X-4-71628393710.1186/1472-684X-4-7Research ArticleThe Australia-modified Karnofsky Performance Status (AKPS) scale: a revised scale for contemporary palliative care clinical practice [ISRCTN81117481] Abernethy Amy P [email protected] Tania [email protected] Belinda S [email protected] David [email protected] David C [email protected] Department of Palliative and Supportive Services, Division of Medicine, Flinders University, Bedford Park, South Australia, Australia2 Southern Adelaide Palliative Services, Repatriation General Hospital, Daw Park, South Australia, Australia3 Division of Medical Oncology, Department of Medicine, Duke University Medical Center, Durham, North Carolina, USA4 North Tasmanian Palliative Care Service, Launceston, Tasmania, Australia2005 12 11 2005 4 7 7 19 5 2005 12 11 2005 Copyright © 2005 Abernethy et al; licensee BioMed Central Ltd.2005Abernethy et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The Karnofsky Performance Status (KPS) is a gold standard scale. The Thorne-modified KPS (TKPS) focuses on community-based care and has been shown to be more relevant to palliative care settings than the original KPS. The Australia-modified KPS (AKPS) blends KPS and TKPS to accommodate any setting of care.
Methods
Performance status was measured using all three scales for palliative care patients enrolled in a randomized controlled trial in South Australia. Care occurred in a range of settings. Survival was defined from enrollment to death.
Results
Ratings were collected at 1600 timepoints for 306 participants. The median score on all scales was 60. KPS and AKPS agreed in 87% of ratings; 79% of disagreements occurred within 1 level on the 11-level scales. KPS and TKPS agreed in 76% of ratings; 85% of disagreements occurred within one level. AKPS and TKPS agreed in 85% of ratings; 87% of disagreements were within one level. Strongest agreement occurred at the highest levels (70–90), with greatest disagreement at lower levels (≤40). Kappa coefficients for agreement were KPS-TKPS 0.71, KPS-AKPS 0.84, and AKPS-TKPS 0.82 (all p < 0.001). Spearman correlations with survival were KPS 0.26, TKPS 0.27 and AKPS 0.26 (all p < 0.001). AKPS was most predictive of survival at the lower range of the scale. All had longitudinal test-retest validity. Face validity was greatest for the AKPS.
Conclusion
The AKPS is a useful modification of the KPS that is more appropriate for clinical settings that include multiple venues of care such as palliative care.
==== Body
Background
Palliative care clinicians are increasingly using change in performance status as a flag for likelihood of need for services, timing of interventions, and as an outcome measurement for clinical programs and research [1-3]. The Karnofsky Performance Scale (KPS) has been used as an assessment tool for performance status in oncology since 1948 [4]. It is commonly regarded as the gold standard measurement of performance status in cancer[2,3]. The KPS scale assesses three dimensions of health status – activity, work and self-care – and can be administered by any healthcare professional for a quick assessment of general functioning and survival [5].
The original KPS is an ordered categorical scale with 11 levels (Table 1). Extensive psychometric testing provides evidence of acceptable reliability and validity in patients with cancer[3,6,7]. The KPS correlates well with physical functioning, such as walking and stair climbing [7]. It has been repeatedly demonstrated to be useful in assisting prognostication [4,8-11]. In an evaluation of predictive validity, Mor et al found significant correlation between KPS at initial interview and survival time (r = 0.30, P < 0.001) [6]. When KPS is low it is a sensitive predictor of poor prognosis, but when high it is a poor cross-sectional indicator of prognosis [2].
Table 1 Comparison of the original Karnofsky Performance Status Scale (KPS), Thorne-modified Karnofsky Performance Status Scale (TKPS), and the Australia-modified Karnofsky Performance Status Scale (AKPS). Italicised areas reflect the original KPS instrument.
Score (Category) Original Karnofsky (KPS) Thorne-modified Karnofsky (TKPS) Australia-modified Karnofsky (AKPS)
100 (A) Normal; no complaints; no evidence of disease. Normal; no complaints; no evidence of disease. Normal; no complaints; no evidence of disease.
90 (A) Able to carry on normal activity; minor signs or symptoms. Able to carry on normal activity; minor signs or symptoms. Able to carry on normal activity; minor signs or symptoms.
80 (A) Normal activity with effort; some signs or symptoms of disease. Normal activity with effort; some signs or symptoms of disease. Normal activity with effort; some signs or symptoms of disease.
70 (B) Cares for self; unable to carry on normal activity or to do active work. Cares for self; unable to carry on normal activity or to do active work. Cares for self; unable to carry on normal activity or to do active work.
60 (B) Requires occasional assistance but is able to care for most of his needs. Requires professional visits less than once a week. Requires occasional assistance but is able to care for most of his needs.
50 (B) Requires considerable assistance and frequent medical care Requires professional visits more than once a week. Requires considerable assistance and frequent medical care
40 (C) Disabled; requires special care and assistance. In bed more than 50% of the time. In bed more than 50% of the time.
30 (C) Severely disabled; hospitalisation necessary; active supportive treatment is necessary. Almost completely bedfast. Almost completely bedfast.
20 (C) Very sick; hospitalisation necessary; active supportive treatment is necessary. Totally bedfast and requiring extensive nursing care by professionals and/or family. Totally bedfast and requiring extensive nursing care by professionals and/or family.
10 (C) Moribund; fatal processes progressing rapidly. Comatose or barely arousable. Comatose or barely arousable.
0 Dead. Dead. Dead.
While useful, the original KPS has limitations. It was based on the models of health service delivery available in 1948, linking performance status with strict recommendations about where further clinical care should be provided (Table 1). At KPS 30 and below there are recommendations for the intensity of clinical care including hospitalization. In particular, the KPS focus on need for hospitalization and medical intervention limits its applicability in care settings where clinical options extend to non-hospital-based care directed at support rather than cure, such as palliative care settings. The language may be uncomfortable or confusing to nurses and other clinicians expected to apply the scale to palliative patients for whom they are caring in home or hospice settings but then expected to blatantly ignore the KPS recommendations of hospitalization. Recommendations about place of care need not be part of the KPS for it to be clinically useful in the 21st century [3].
These limitations have prompted modifications of the scale and development of new tools that better reflect clinical functioning and variations in place of care for palliative care patients. In the late 1990's, Thorne developed a modified version of the KPS (Thorne-modified Karnofsky Performance Status, TKPS, Table 1) for use in palliative home care settings. The TKPS reworded the categories at the lower end of the KPS scale to correlate with professional care needs and activity, removing references to location of care. The TKPS was validated by Nikoletti and colleagues in a sample of 78 Australian home-hospice patients [3].
While the TKPS was found to be more applicable for home-based hospice settings, it was limited in its use for hospitalized palliative care patients. In response, the original KPS and the newer TKPS were melded into a single scale that accommodated all of the venues of clinical palliative care – the Australia-modified Karnofsky Performance Status Scale (AKPS, Table 1). In the AKPS, the TKPS link to health professionals' visits was avoided, favoring a generic approach focusing on function alone.
We conducted a large randomized controlled trial that incorporated performance status as a main outcome measure – the Palliative Care Trial. This study is a 2 × 2 × 2 factorial cluster randomized controlled trial involving 461 consenting patients and their general practitioners (GPs) recruited from April 2002 and June 2004. Participants were randomized to case conferences, GP pain education, and patient pain education. Patient participants were cared for in a variety of clinical settings consistent with contemporary clinical practice. It was important that the performance status scale used was appropriate and carefully validated. The full clinical trial methodology has been presented elsewhere [1].
The purpose of this current study was to determine the performance status measure most appropriate for the clinical and research needs of the community based palliative care service conducting the Palliative Care Trial. The expected goals were to do the following:
1. To determine if the TKPS and AKPS had similar predictive values for survival as compared with the KPS in a contemporary specialized palliative care service that incorporates a range of clinical settings including acute hospital care, inpatient hospice, community care, and aged care facilities;
2. To determine if one version of the KPS instrument was clearly superior to the others in this setting;
3. If equal, to determine which instrument was most acceptable to clinical and research staff and easiest to use; and,
4. To document reliability of the three instruments in the local clinical setting.
Methods
This current study was an embedded sub-study within the Palliative Care Trial [1]. It was decided a priori that after KPS, TKPS and AKPS data were collected from at least 120 participants who exited the trial, a validity assessment of the scales would be conducted. Based on this analysis all subsequent Palliative Care Trial assessments would include only the most reliable of these 3 measures.
Study setting
The trial was set in Adelaide, South Australia and based at Southern Adelaide Palliative Services. GPs are the primary point of care for palliative patients. Specialized palliative care services funded by the state government provide consultative specialist medical and nursing support for GPs and community nurses. Referrals to the palliative care service come from health professionals, family and patients, and nearly all palliative care patients within the region are referred to the same geographic service. The indication for, and timing of referral varies; indications include physical symptom control, and the need for coordinated and multi-disciplinary care. Patients frequently receive active therapy such as radiotherapy and chemotherapy in parallel with palliative care services. The model of care is consistent with the definition of palliative care described in the 2004 United States (US) National Consensus Project Clinical Practice Guidelines for Quality Palliative Care [12].
Southern Adelaide Palliative Services is a comprehensive program with medical specialists, nursing specialists, social work, inpatient care, community and outpatient visits, home care, nursing home consultations, a bereavement program, volunteers, and complementary care. Venues of care include community, acute inpatient, sub-acute inpatient, inpatient hospice, respite, nursing home, and hostel settings. There were 1,094 new referrals in 2003, 85% of whom had cancer. The mean time from referral to death was 119 days with a median of 47 days. Ten percent of referrals were from GPs, 50% from local hospitals, 20% from medical specialists, and 10% from district nurses. Ninety percent of patients spent some time in the hospital in the last year of life and less than one third spent any time in the inpatient hospice.
Study participants
All adult patients referred to Southern Adelaide Palliative Services with any form of pain in the preceding three months were eligible for the Palliative Care Trial. This definition was very broadly applied and could refer to any type of pain. This pain could have been temporary and not related to the predominate illness, but would provide a reference point for pain assessments during the research study. Patients who did not live within the geographic region served by the palliative care service were excluded. Patients who were expected to die within 48 hours of referral were also excluded, since the recruitment process of consenting both the patient and GP was expected to take two days. Participants must have been mentally competent at enrollment as documented by a Folstein Mini-mental Status Examination (MMSE) Score = 24 [13], or have a GP-identified caregiver or legal healthcare proxy who could adequately provide informed consent [14]. Patients unwilling to provide their contact information to the trial staff in order to learn more about the study were not enrolled. All GPs and GP practices of consenting eligible patients were eligible. Both patient and GP consent were required for enrollment and randomization.
Measures
Patient participant functional status was assessed by the KPS, TKPS and AKPS (Table 1) simultaneously. For those patients who died during the trial, survival was defined as time from consent to participate in the trial until death.
Study procedures
KPS, TKPS and AKPS were collected at every clinical encounter and data collection time point (minimally baseline and every 2 weeks for the first 12 weeks then monthly until death or exit from the trial) for the first 300 participants randomized in the trial. Based on our sample size assumptions and recruitment estimates, this would provide data for over 120 participants who exited the trial through either death or withdrawal from the study. Palliative care clinical nurses collected all of the data within the study. Data collection was incorporated into regular palliative clinical visits irrespective of location of care (hospital- or community-based care). In total, 26 full or part-time nurses collected data.
All clinical nurses participated in training in performance status assessment. This included at least a 30-minute review of the scales, data collection forms, and reasons for using all three measures. In addition, written instructions were prepared that included a list of example questions that could be used to determine performance status classification. Nurses were provided laminated cards with each of the scales. Responses were captured on a standardized set of study forms. Nurses also underwent more specific training on data collection methods, the importance of data quality, communication and research ethics as part of the Palliative Care Trial [1].
Analysis
Descriptive statistics were used to summarize the participant population and performance status scores observed. Agreement between the KPS, TKPS and AKPS was assessed by a variety of methods. Percentage agreements and disagreements were calculated; overall agreement and proportion of disagreements at one, two or three levels on the scales were reported in the method of Nikoletti et al [3]. Kappa statistics were calculated in order to exclude agreement due to chance. Simple kappa statistics were reported and subjected to the following interpretation: 0.81 to 1.00 almost perfect agreement, 0.61 to 0.80 substantive agreement, 0.41–0.60 moderate agreement, 0.21–0.40 fair agreement, 0.00–0.20 slight agreement, and less than 0.00 poor or no agreement [15]. Bland-Altman plots of the difference between paired scores versus the mean of the scores were used to assess agreement [16]. The association between performance status and survival was determined using Spearman correlation coefficients [5]. Actuarial survival was estimated using Kaplan-Meier methods and reported as median (interquartile range (IQR)) survival [17]. Between group comparisons were made using the log rank test [18]. All analyses were conducted with the SAS System (Version 9.1, Cary, North Carolina, USA).
Ethics approval and trial registration
The Palliative Care Trial was approved by all twelve relevant independent Institutional Review Boards (IRBs) and Human Research and Ethics Committees (HRECs) including the Australian Department of Veterans Affairs and Health Insurance Commission, Canberra, Australia. This sub-study was included in the approved protocols and participant consent forms. The Palliative Care Trial is registered with the ISRCTN – ISRCTN81117481 .
Results
KPS, TKPS and AKPS were assessed simultaneously 1600 times in a total of 306 patients. There were a mean of 5.3 (standard deviation (SD) 5.0) assessments per individual, with median 4.0 (range 1–44). Within the same individual, the multiple performance assessments were collected over a mean of 76 (SD 88) days, with median 49 (range 0–507). Seventy-eight percent of assessments were done in the patient's home, 7% in an aged care facility, 5% in another relative's home, 5% in the hospital, 4% in the inpatient hospice unit, and 1% in another location of care.
The baseline characteristics of the patient population are presented in Table 2. The mean age of the study population was 71 years, 148 (49%) were male, 190 (63%) married, and 69 (23%) widowed. The majority (93%) had cancer and had a primary caregiver (96%). Fourteen patients had a MMSE score of <24; of those, 10 had an AKPS of <60.
Table 2 Baseline participant characteristics
Characteristic Category N %
Gender Male 148 49%
Age (Mean/SD) 71 12%
Marital status Never Married 9 3%
Widowed 69 23%
Divorced/Separated 33 11%
Married/Defacto 190 63%
Caregiver present Has caregiver 277 96%
Accommodation Private residence 271 89%
Nursing home 18 6%
Hostel 4 1%
Hospital 11 4%
Other 2 1%
Living arrangement Lives alone 72 25%
Lives with spouse 173 60%
Other relative lives in household 44 15%
Cancer diagnosis Yes 282 92%
Phase of palliative care Stable 157 60%
Deteriorating 48 18%
Unstable 57 22%
Terminal 1 0%
Pain at present Mean (standard deviation) (2.1)
Usual pain in last 24 hrs (2.1)
Worst pain in last 24 hrs (3.2)
The profile of KPS, TKPS and AKPS scores in is shown in Figure 1. The majority of scores centered around 50–70 (group B). Measures of central of central tendency are presented in Table 3; the median score was 60 on all three scales.
Table 3 Measures of central tendency and dispersion for the KPS, TKPS and AKPS scores
N Mean Median Mode SD Min-Max
KPS 1600 59.7 60 50 15.7 10–100
TKPS 1600 59.0 60 70 17.3 10–100
AKPS 1600 58.7 60 70 17.1 10–100
Figure 1 Profile of KPS, TKPS and AKPS scores in the 1600 observations.
The agreement between scores was calculated for KPS-TKPS, KPS-AKPS and TKPS-AKPS pairings, as shown in Table 4; relevant scatter plots are in Figure 2. These demonstrate a high level of agreement between the 3 scales. When disagreement existed, KPS scores were more commonly higher than the TKPS or AKPS scores; the profile of disagreement between TKPS and AKPS was more evenly split and nearly always confined to one level. Bland-Altman plot presented in figure 3 verify these interpretations.
Table 4 Comparison between KPS, TKPS and AKPS scores by levels
N %
Comparison between scores for KPS and TKPS
By score
TKPS > KPS by three levels 1 0.1%
TKPS > KPS by two levels 7 0.4%
TKPS > KPS by one levels 149 9.3%
Complete agreement 1216 76.0%
TKPS < KPS by one levels 179 11.2%
TKPS < KPS by two levels 48 3.0%
TKPS < KPS by three levels 0 0.0%
By Group (ABC)
TKPS >KPS by one group 96 6.0%
Complete agreement 1489 93.1%
TKPS <KPS by one group 15 0.9%
Comparison between scores for KPS and AKPS
By score
AKPS > KPS by three levels 1 0.1%
AKPS > KPS by two levels 1 0.1%
AKPS > KPS by one levels 44 2.8%
Complete agreement 1386 86.6%
AKPS < KPS by one levels 123 7.7%
AKPS < KPS by two levels 45 2.8%
AKPS < KPS by three levels 0 0.0%
By Group (ABC)
AKPS >KPS by one group 80 5.0%
Complete agreement 1507 94.2%
AKPS <KPS by one group 13 0.8%
Comparison between scores for TKPS and AKPS
By score
TKPS > AKPS by three levels 0 0.0%
TKPS > AKPS by two levels 8 0.5%
TKPS > AKPS by one levels 139 8.7%
Complete agreement 1359 84.9%
TKPS < AKPS by one levels 88 5.5%
TKPS < AKPS by two levels 6 0.4%
TKPS < AKPS by three levels 0 0.0%
By Group (ABC)
TKPS >AKPS by one group 15 0.9%
Complete agreement 1556 97.3%
TKPS <AKPS by one group 29 1.8%
Figure 2 Scatter plots of KPS-TKPS, KPS-AKPS, and AKPS-TKPS pairs.
Figure 3 Bland & Altman plot for the performance status pairs.
Disagreements tended to cluster at the mid-point and lower potions of the scales as shown in the scatter plots of Figure 2. The Kappa coefficient for agreement between all KPS and TKPS measurements was 0.71 (p < 0.001), between all KPS and AKPS measurements was 0.84 (p < 0.001), and between all AKPS and TKPS measurements was 0.82 (p < 0.001).
All three performance measurements were correlated with survival. For those participants who died while on the trial (n = 232), the Spearman correlation coefficients of baseline KPS, TKPS, and AKPS and overall survival were 0.26, 0.27, and 0.26 respectively all with p < 0.001. These positive correlations imply that when performance status increased, survival also increased. Since palliative patients are more likely to have poor performance status, the predictive power of the three instruments in the intermediate (category B) and lower (category C) ranges was investigated (Figures 4 and 5). For category B, all three instruments clearly discriminated survival for levels 50, 60 and 70 (p < 0.001; Figure 4). For category C, only AKPS could significantly discriminate survival based upon AKPS levels of 30 and 40 (p = 0.026; Figure 5).
Figure 4 Survival probabilities according to KPS, TKPS and KPS by levels 50, 60 and 70 (Category B).
Figure 5 Survival probabilities according to KPS, TKPS and KPS by levels 30 and 40 (Category C).
Longitudinal test-retest reliability indicates the ability to have multiple longitudinal assessments that are sensitive to meaningful change in the outcome. Since we would be measuring performance status over time in the trial, this was of considerable concern for us. KPS, TKPS, and AKPS longitudinal curves were generated for all participants; an example is given in Figure 6. Longitudinal trends consistent with the established palliative care trajectories of illness were observed with all instruments (data not shown). Change in performance status level was investigated for each participant when moving from the stable to the deteriorating phase of palliative care (n = 96). Palliative care phase is predictive of the need for palliative care service intervention and resource utilization [19,20]. When phase deteriorated, KPS, TKPS and AKPS all decreased by a median of 10 (range -10 to 50).
Figure 6 Demonstration of the ability of the three instruments to respond to change in performance status over time. Plots nearly overly each other. Similar plots were generated for all participants.
Face validity was assessed by asking nurses to indicate which scale was easiest to use and best suited the patient population. Nurses involved in the collection of data for the trial were highly skilled and experienced in palliative care with a mean length of time in nursing of 18 years (range 6–29) with an average of 9 years (range 2–16) in palliative care. All nurses had university qualifications; 24 of the nurses were employed at RN level 2 (consultant) and 2 employed at RN Level 3 (consultant and managerial). All 26 nurses preferred using the AKPS as they felt that the categories were more consistent with the multiple venues of care and different levels of interventions needed in their current clinical practice.
Discussion
KPS is a widely used measure demonstrated to have important correlations with both resource utilization and prognosis at the end of life. Generating a measure of function with language that is consistent with current clinical practice is crucial for the ongoing use of KPS. For the measure to transcend differences in funding of healthcare and models of service delivery, it is timely to remove specific references to the place or intensity of care and focus on function alone.
The Thorne modification developed in the 1990's was an important validated update, making the scale useful for contemporary palliative home care settings, especially hospice. The TKPS concentrated on the community setting, though, limiting the scale's utility in the varied clinical settings encountered in palliative care including inpatient hospice, acute inpatient care, and nursing home care. This is the first research report of the Australia-modified version (AKPS), an important amalgam of the original KPS and the TKPS applicable to both inpatient and community palliative care. The categories in the AKPS are less directive of the expected location of care; however, as much of the original KPS and TKPS language as possible has been maintained in order to reduce confusion and the need for extensive retraining for clinicians already familiar with the earlier versions.
In this study all versions of the KPS could be used in the various venues of palliative care including the community, acute inpatient, subacute inpatient, inpatient hospice, respite, nursing home, and hostel settings. AKPS had the highest agreement with both KPS and TKS (Table 4, Figures 2 and 3), and was equally predictive of survival (Figure 4). When considering the lower end of the scale (category C) where more palliative patients cluster, AKPS was most predictive of survival (Figure 5). All scales were able to reflect longitudinal change. The nurses reflected that AKPS was easiest to use and most acceptable. This study demonstrated that AKPS had excellent correlation with the original KPS while allowing for palliative care sensitive clinician responses to changes in level of function as death approaches, both in terms of place of care and the clinical staff who need to be involved in that care. The better performance of the AKPS will assist with better decision-making in palliative care.
The high level of agreement among the three versions was expected, given the similarity of the three scales. However, before a new scale is adopted for day-to-day clinical practice it is important that it is carefully and prospectively evaluated to ensure that the results reflect what the user expects to be measuring. Further, as we planned to use performance status as a primary outcome in a major clinical trial in palliative care it was vital to verify the validity of the AKPS as an outcome measure within the palliative care setting before limiting all of our data collection to this single measure. The need for formal validation is evident in Table 4 and Figure 2. Some participants were assessed as a KPS of 20 and an AKPS of 50 even when the KPS 50 and AKPS 50 had exactly the same phrasing. This was done by the same nurse assessing the patient using each of the scales sequentially at the same evaluation visit. This difference in scoring was reflective of the difference in phrasing at other levels on the scales. For an individual palliative care patient, a score of KPS 20 ("very sick; hospitalization necessary; active supportive treatment necessary") may be the best option on that scale, however when reviewing the AKPS scale the score of 50 ("requires considerable assistance and frequent medical care") was more appropriate in relation to all other levels on the scale. Importantly, the AKPS instrument with more palliative care appropriate language did not alter the scale's expected overall correlation with survival, was more correlated with survival at lower performance status, and was more acceptable to the clinical nurses.
Longitudinal assessment
Any of the KPS tools provide both an objective measure of the current status of the person being assessed and useful trends when used longitudinally. Longitudinal trends in performance status is an important aspect of prognostication for the longer term outlook of the patient including his or her anticipated health resource and service needs over time, as demonstrated by the relationship between performance status and phase of palliative care [20]. Such changes in level of function reflect the disease trajectories described by Lunney, Lynn and colleagues [21], irrespective of the underlying life-limiting illness.
Other measures of performance
KPS is not the only measure of performance status used in palliative care and oncology. The shorter Eastern Cooperative Oncology Group (ECOG) performance status scale was derived from the KPS [22]. It is only occasionally used as a main outcome in clinical trials in the palliative care setting since the 5-item scale inadequately differentiates between patients with poor functional status. Similar concerns about the KPS having limited sensitivity to monitor change when patients score at the low end of the scale have been reported by other authors [2]. In this current study, AKPS was superior to KPS and TKPS in the lower range of the scale and provided more categorical levels of performance status than the ECOG scale.
In 1996, Anderson et al described the Palliative Performance Scale (PPS), a modification of the KPS based on 5 observable parameters (ambulation, activity combined with evidence of disease, self-care, intake and conscious level) scored into 11 categories. [23]. PPS predicted time to death (mean 162 days) for a population of Australian patients admitted to a palliative care unit[24]. PPS was not as predictive of survival for a similar group of Japanese palliative care patients with a mean survival of 49 days [11]. AKPS is a less complex measure which is easier to use with each clinical encounter.
Other scales such as the Edmonton Functional Assessment Tool extend on the functional parameters described in the KPS, PPS and TKPS [2], however as these scales and their scoring becomes more complicated, their day-to-day applicability decreases. AKPS focuses on current functional abilities and on changes in function if used longitudinally; it provides an important parameter in the overall assessment of any person with a life-limiting illness.
Limitations
This study is representative by age and gender for palliative care in Australia. Because it was a sub-study of a larger randomized controlled trial where pain in the previous 3 months was an inclusion criterion, the population almost all had cancer as their life-limiting illness (92%) versus 85% seen in the general population referred to the palliative care service. Given that KPS was originally developed for people with cancer and had been extrapolated to other clinical settings (AIDS, end-stage organ failure), this should not be a major limit to generalizing these findings. In the early parts of the trial, measurement of KPS, TKPS and AKPS was predominantly in the community setting limiting the ability to observe its utility in other care venues. As more trial participants were hospitalized over time, data were collected from the inpatient settings therefore reflecting the range of settings in which palliative care is delivered. Many assessments were in the upper range of the scales; only a minority of patients were bedridden. An evaluation on a palliative care unit with more severely disabled patients might show other results. The missing correlation of KPS and TKPS with survival in the lower range of the scale may have been biased by small patient numbers in these clusters. Also, ideally none of the performance status measures would have any reference to the amount of health services required at any of the levels. AKPS has considerably less reference, but still states "requires... frequent medical care" in its description of AKPS 50.
Inter-rater reliability evaluation of the AKPS was originally planned as part of this sub-study. Ill palliative care patients were overly burdened by multiple visits on the same day for research data collection. An alternative plan was enacted with collection of measures after the informed consent document was signed and then comparing these results with those reported on the baseline assessment within 48 hours of the consent visit. Unfortunately many patients were too unstable or the timing of the baseline assessment was too far from the consent visit; there were not enough data available for these analyses. The inter-rater reliability and other psychometric properties of the KPS has previously been documented[3,6,7]. As AKPS was more predictive of survival outcomes than KPS, the reliability was expected to be better if it differed from that reported for the KPS.
In addition to being more predictive of survival at the lower end of the scale, the AKPS may be more appropriate than the other performance status scales in settings outside of cancer. Ninety-two percent of participants in this study had cancer, so evaluation of the AKPS in non-cancer diagnoses was limited. Further work should concentrate on disabled palliative care patients whose performance status is at the lower end of the scale. Future studies will focus on validity outside of the cancer setting and with more diverse palliative care populations. Also, the performance of the AKPS will be compared to other performance status measurement tools appropriate for palliative care such as the ECOG scale, PPS and Edmonton Functional Assessment Tool.
Conclusion
The AKPS is an important contemporary modification to the KPS that incorporates language more appropriate to current clinical care without limiting it to judgments on intensity of clinical treatment or resources available. The AKPS assists with clinical decision-making across a range of clinical settings. The ability of staff to relate to the performance status scale used will continue to be a key factor in uptake and use of a measure that has such broad application. It is appropriate to use the AKPS as a primary outcome variable in the Palliative Care Trial.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
APA: Conception and design, data preparation, data analysis, analysis interpretation, preparation of the first and subsequent manuscript drafts, and final approval of the manuscript
T S-J: Conception and design, data collection, data preparation, data analysis, analysis interpretation, review of manuscript drafts, and final approval of the manuscript
BSF: Conception and design, analysis interpretation, review of manuscript drafts, and final approval of the manuscript
DW: Developed the AKPS, approval of design of the validation study, analysis interpretation, review of manuscript drafts, and final approval of the manuscript
DCC: Conception and design, analysis interpretation, preparation of the first and subsequent manuscript drafts, and final approval of the manuscript
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors thank Dr. David Thorne, the developer of the TKPS. We also thank Dr. Greg Samsa, PhD statistician, for his advice on the analysis plan. Many people have worked together selflessly and enthusiastically to make the Palliative Care Trial a success; a complete list of these individuals is provided elsewhere[1]. In particular, the authors would like to thank the nurses of Southern Adelaide Palliative Care Services (Adelaide, South Australia) who collected the data for this study. We also thank all of the study participants who generously donated their time and personal information in an effort to improve the care of others who need palliative services.
Direct costs of the Palliative Care Trial were provided through a grant from the Rural Health and Palliative Care Branch of the Australian Department of Health and Ageing (Canberra, Australia), under the National Palliative Care Strategy. Additional funds for the educational intervention and its evaluation were provided by the Ian Potter Foundation (Melbourne, Australia), Cancer Council South Australia (formerly the Anti-Cancer Foundation; Adelaide, Australia), and the Doris Duke Charitable Foundation (New York City, USA). Dr. Abernethy's salary is generously provided through a Clinical Scientist Development Award from the Doris Duke Charitable Foundation of New York, New York, USA. The study was jointly sponsored by the Australian Department of Health and Ageing (Canberra, Australia), Repatriation General Hospital (Daw Park, South Australia), ACH Group, Inc. (Adelaide, South Australia), and Southern Division of General Practice (Bedford Park, South Australia). The design, conduct, analysis and write-up of the study were performed independently from any funding or sponsoring agency.
==== Refs
Abernethy AP Currow DC Hunt R Williams H Roder-Allen G Rowett D Shelby-James T Esterman A May F Phillips PA A pragmatic 2x2x2 factorial cluster randomized controlled trial of educaitonal outreach visiting and case conferencing in palliative care. Methodology of the Palliative Care Trial Contemp Clin Trials 2005 16290094
Kaasa T Wessel J The Edmonton Functional Assessment Tool: further development and validation for use in palliative care Journal of Palliative Care 2001 17 5 11 11324186
Nikoletti S Porock D Kristjanson LJ Medigovich K Pedler P Smith M Performance status assessment in home hospice patients using a modified form of the Karnofsky Performance Status scale Journal of Palliative Medicine 2000 3 301 311 15859671 10.1089/jpm.2000.3.301
Karnofsky D Abelmann W Craver L Burchenal J The use of nitrogen mustard in the palliative treatment of cancer. Cancer 1948 1 634 656
Buccheri G Ferrigno D Tamburini M Karnofsky and ECOG performance status scoring in lung cancer: A prospective, longitudinal study of 536 patients from a single institution European Journal of Cancer Part A 1996 32 1135 1141 8758243 10.1016/0959-8049(95)00664-8
Mor V Laliberte L Morris JN Wiemann M The Karnofsky Performance Status Scale. An examination of its reliability and validity in a research setting Cancer 2002 53 2002 2007 6704925
Yates JW Chalmer B McKegney FP Evaluation of patients with advanced cancer using the Karnofsky performance status Cancer 1980 45 2220 2224 7370963
Bennett M Ryall N Using the modified Barthel index to estimate survival in cancer patients in hospice: observational study BMJ December 2, 2000 321 1381 1382 10.1136/bmj.321.7273.1381
Evans C McCarthy M Prognostic uncertainty in terminal care: Can the Karnofsky index help? Lancet 1985 1 1204 1206 2860399 10.1016/S0140-6736(85)92876-4
Maltoni M Nanni O Pirovano M Scarpi E Indelli M Martini C Monti M Arnoldi E Piva L Ravaioli A Cruciani G Labianca R Amadori D Successful validation of the palliative prognostic score in terminally ill cancer patients Journal of Pain & Symptom Management 1999 17 240 247 10203876 10.1016/S0885-3924(98)00146-8
Morita T Tsunoda J Inoue S Chihara S Improved accuracy of physicians' survival prediction for terminally ill cancer patients using the Palliative Prognostic Index Palliative Medicine 2001 15 419 424 11591094 10.1191/026921601680419474
American Academy of Hospice and Palliative Medicine Care CAP Association HPN Last AP National Hospice and Palliative Care Organization National Consensus Project for Quality Palliative Care: Clinical Practice Guidelines for quality palliative care, executive summary Journal of Palliative Medicine 2004 7 611 627 15588352 10.1089/jpm.2004.7.611
Folstein MF Folstein SE McHugh PR "Mini-mental state". A practical method for grading the cognitive state of patients for the clinician Journal of Psychiatric Research 1975 12 189 198 1202204 10.1016/0022-3956(75)90026-6
Masso M Dodds S Fildes D al Ethical research in palliative care: a guide through the Human Research Ethics Committee process 2004 Canberra, Australia, Commonwealth of Australia Department of Health and Ageing 1 31 15560112
Cohen J A coefficient of agreement for nominal scales Educ Psychol Meas 1960 20 37 46
Bland JM Altman DG Comparing methods of measurement: why plotting difference against standard method is misleading Lancet 1995 346 1085 7 7564793 10.1016/S0140-6736(95)91748-9
Kaplan EL Meier F Non-parametric estimation from incomplete observations. J Am Stat Assoc 1958 58 457 481
Peto R Pike M Armitage P Design and analysis of randomized clinical trails requiring prolonged observation of each patient: II Analysis and examples. Br J Cancer 1977 35 39
Eagar K The Australian National Sub-Acute and Non-Acute Patient casemix classification Australian Health Review 22(3):180-96, 1999
Eagar K Green J Gordon R An Australian casemix classification for palliative care: technical development and results Palliative Medicine 2004 18 217 226 15198134 10.1191/0269216304pm875oa
Lunney JR Lynn J Foley DJ Lipson S Guralnik JM Patterns of functional decline at the end of life Jama 2003 289 2387 2392 12746362 10.1001/jama.289.18.2387
Michael M Tannock IF Measuring health-related quality of life in clinical trials that evaluate the role of chemotherapy in cancer treatment. CMAJ Canadian Medical Association Journal 1998 158 1727 1734
Anderson F Downing GM Hill J Casorso L Lerch N Palliative performance scale (PPS): a new tool Journal of Palliative Care 1996 12 5 11 8857241
Virik K Glare P Validation of the palliative performance scale for inpatients admitted to a palliative care unit in Sydney, Australia Journal of Pain & Symptom Management 2002 23 455 7 12067769 10.1016/S0885-3924(02)00407-4
|
16283937
|
PMC1308820
|
CC BY
|
2021-01-04 16:30:52
|
no
|
BMC Palliat Care. 2005 Nov 12; 4:7
|
utf-8
|
BMC Palliat Care
| 2,005 |
10.1186/1472-684X-4-7
|
oa_comm
|
==== Front
BMC PediatrBMC Pediatrics1471-2431BioMed Central London 1471-2431-5-411628394610.1186/1471-2431-5-41Research ArticleCase-control study of sudden infant death syndrome in Lithuania, 1997–2000 Bubnaitienė Vilija [email protected]ėdienė Ramunė [email protected]ėvalas Rimantas [email protected] Department of Pediatrics, Kaunas University of Medicine, Eiveniu 2, 5009 Kaunas, Lithuania2 Department of Public Health, Kaunas University of Medicine, Eiveniu 4, 5009 Kaunas, Lithuania2005 13 11 2005 5 41 41 23 1 2005 13 11 2005 Copyright © 2005 Bubnaitienė et al; licensee BioMed Central Ltd.2005Bubnaitienė et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
To identify risk factors for sudden infant death syndrome relevant in Lithuania.
Methods
A nationwide case-control study surveying parents of 35 infants who died from sudden infant death syndrome during the period of 1997–2000 and parents of 145 control infants matched with SIDS infants for date of birth and for region of birth was carried out.
Results
Deaths incidence was greater in the warm period (60%) vs. cold period (40%). Prone and side sleeping positions both carried no increased risk of sudden infant death syndrome compared with supine because of a rare prone sleeping (4.1% of controls vs. 0% of dead infants) and more prevalent side than supine sleeping (84.8% of controls vs. 94.3% of dead infants) in the controls as well as the cases. Bed sharing for the whole night as a risk factor for sudden infant death syndrome has not been confirmed, either, as bed sharing was common only for the controls (13.8% of controls vs. 0% of dead infants). Routine sleeping environment factors such as heavy wrapping (≥4 togs) of an infant (odds ratio 8.49; 95% confidence interval 2.38 to 30.32), sleeping in a bassinet (4.22; 1.16 to 15.38) and maternal factors such as maternal education ≤12 years (4.48; 1.34 to 14.94), unplanned pregnancy (5.22; 1.49 to 18.18) and ≥2 previous live births (3.90; 1.00 to 15.10) were significantly associated with sudden infant death syndrome on multivariate analysis.
Conclusion
The results of this first population-based case-control study have shed some light on the epidemiology of the syndrome in Lithuania. Although the mortality of sudden infant death syndrome in Lithuania is not high, it might be lowered moreover by public informing about sudden infant death syndrome and related risk factors. Special attention must be paid to mothers with low education on potentially modifiable risk factors such as routine heavy wrapping of an infant during sleep, routine sleeping in a bassinet and unplanned pregnancy.
==== Body
Background
Despite the fact that sudden infant death syndrome (SIDS) is the leading cause of postneonatal infant mortality in most developed countries, SIDS incidence varies greatly in different countries and between regions within countries worldwide. Before 1990–1991, SIDS incidence varied from 1 to 6 cases per 1000 live births [1]. Since 1991 SIDS incidence has declined significantly in a lot of countries and now varies from 0.1 to 1.5 cases per 1000 live births [2]. Unfortunately, the cause of SIDS and variations of incidence in different countries remain unclear. In recent years, there has been considerable interest in the role of infant care practices and sleeping environment in SIDS. Sleeping prone has been found to be an especially strong and consistent risk factor across different societies and countries, and modification of this practice has been associated with a major reduction in SIDS incidence [3,4]. In the meantime the average mortality rate from SIDS in Lithuania during the period of 1997–2000 was 0.3 per 1000 live births and was low, if compared to international mortality rates though no risks reducing campaign has been performed.
As considerable differences of SIDS risk factors importance may exist in different countries, the aim of our study was to identify factors associated with and predicting increased risk of SIDS relevant in Lithuania.
Materials and methods
Study design and subjects
The survey was performed as a retrospective case-control study during the period of 2002–2003. The main case-control study instruments were questionnaires for cases and controls. The questionnaire for cases consisted of 89 standardized questions concerning infant death circumstances, demographic factors, routine practices in sleeping environment, infant and maternal medical history, parental socioeconomic status and lifestyle. The questionnaire for controls consisted of 81 standardized questions concerning demographic factors, routine practices in sleeping environment, infant and maternal medical history, parental socioeconomic status and lifestyle. No reference sleep was assigned for the control group as the time frame was too long. Questions were of a "yes" or "no" or multiple choice nature or needed data in figures, such as birth date, birth weight, death time, and others.
Questionnaires for cases were completed by the research interviewer during a home visit. Dialogue about aims of visit took the priority. Afterward standardized questions for SIDS cases were asked. SIDS parents were questioned only on a receipt of underwritten consent of participation in the study. The mean time between SIDS death and completing the questionnaire was 3.9 ± 0.2 years. Questionnaires for controls together with a sheet of information and consent of participation in the study were distributed by mail to control parents.
Data of SIDS infants were obtained from computerized database of Lithuanian Department of Statistics. Initially all of 45 infants whose death was attributed to sudden infant death syndrome (ICD-10 – R 95 and R 96) during the period of 1997–2000 were included in the study. Data of control infants were obtained from Lithuanian Medical Birth Registry. Initially 225 controls matched with SIDS infants for date of birth within one month and for region of birth were picked randomly. Selection results were rectified according to Lithuanian Infant Mortality Registry data and 3 infants who died during the period of 1997–2000 were excluded from the control sample.
Statistical analysis
Data of the study were processed using STATISTIKA/w 5 and SPSS/w 10 (Statistical Package for Social Science) software.
Student t and Fisher's (exact) criteria were used for comparison of means. As using these criteria is possible only when sample variables are distributed normally, Kolmogorov-Smirnov goodness-of-fit test was performed in order to assess the normality of distribution.
Chi-square and logistic regression analyses were used to examine differences in the prevalence and prediction of various sleeping environment, maternal, infant, parental socioeconomic and lifestyle risk factors among SIDS cases and controls. Fisher's exact test was utilized for small cell frequencies. Variables significant on univariate analysis were included in multivariate analysis using a conditional logistic regression. Later on, multivariate models were constructed with the backward stepwise procedure for variables significant at the 5% level. The estimations of both univariate and multivariate analysis resulted in odds ratio (OR) with their 95% confidence intervals (CI).
The hypotheses were considered statistically significant at the level of p < 0.05, very significant at the level of p < 0.01 and especially significant at the level of p < 0.001. The hypotheses were considered statistically not significant at the level of p > 0.05.
Results
Response cases and controls
10 SIDS cases (22.2%) of the total 45 were excluded from the analysis: 5 because the families could not be traced, having moved from the notified inhabitation place; 2 because the families refused an interview and 3 because they fell short of accepted SIDS definition criteria after review of the case history on the grounds of an interview. We defined SIDS according to Willinger as the sudden death of an infant aged from 7 to 365 days, which remained unexplained after performance of a complete postmortem investigation, including an autopsy, examination of the death scene, and review of the case history [5]. Of the potential 222 control families 40 were not available, 20 refused an interview and 17 were rejected.
Finally, data of 35 SIDS cases and 145 controls (1:4) have entered the final statistical analysis.
Deaths of sudden infant death syndrome: relation to age, season and time of death
The mean (± SE) age at death in SIDS infants was 114.1 ± 9.6 days and varied from 23 to 235 days. A χ2 test of the null hypothesis of a uniform frequency distribution over 12 months was significant (χ2 = 15.5; df = 11; p <0.05). The proportion of SIDS deaths was greater in the first half-year of age (82.8%) than in the second half-year (17.2%) with a peak incidence from 2 to 4 month of age (59.9%) (Figure 1).
Figure 1 Age at death in months of SIDS infants (n = 35).
SIDS deaths according to season were classified into two periods – cold period (six lowest temperature months – October to March) and warm period (six highest temperature months – April to September) (Figure 2). SIDS deaths incidence was greater in the warm period (60%) vs. cold period (40%).
Figure 2 Seasonal distribution of SIDS infants (n = 35).
Most of the SIDS deaths occurred during what the parents classified as night sleep. 74.3% of infants (26/35) were discovered dead between 1.50 and 9.00 hours and 25.7% (9/35) – between 9.30 and 14.30 hours. The interval between the time that infants were last seen or heard alive and the time found dead was 186.2 ± 24.6 min.
Univariate results
A univariate analysis was carried out to estimate the strength of the relationship between SIDS and factors concerning routine practices in sleeping environment, infant and maternal medical history, parental socioeconomic status and lifestyle. Within view of SIDS mothers reluctance to report alcohol consumption, especially frequency of alcohol consumption, this variable had not been involved into univariate analysis. Frequencies, odds ratios and p values of all the univariate findings are presented in Table 1. Fisher's exact test was utilized for variables with small cell frequencies. The baseline comparison group always had the opposite definition, for example, infant birth weight <2500.0 g was compared with infant birth weight ≥2500.0 g, unless otherwise stated in the table.
Table 1 Univariate findings of sudden infant death syndrome
Variables Proportion (%) of: Odds ratio (95% CI) p
Cases Controls
SLEEPING ENVIRONMENT FACTORS*
Position when put down to sleep:
• supine 2/35 (5.7) 16/145 (11.0) 1.00 (reference)
• side 33/35 (94.3) 123/145 (84.8) 2.15 [0.47; 9.80] 0.325
• prone 0/35 (0) 6/145 (4.1) 0.00 0.999
Sleeping surface:
• crib 15/35 (42.9) 113/145 (77.9) 1.00 (reference)
• bassinet 19/35 (54.3) 16/145 (11.0) 8.95 [3.80; 21.05] <0.001
• full-sized bed (parental or sofa) 1/35 (2.9) 16/145 (11.0) 0.47 [0.05; 3.81] 0.480
Substandard infant mattress** 25/35 (71.4) 66/145 (45.5) 3.09 [1.38; 6.89] 0.006
Waterproof cloth over the mattress 18/35 (51.4) 45/145 (31.0) 2.35 [1.11; 4.98] 0.025
Use of a pillow# 32/35 (91.4) 135/145 (93.1) n/a 0.268
Wearing ≥4 togs 24/35 (68.6) 31/145 (21.4) 8.02 [3.55; 18.16] <0.001
Wearing a cap 17/35 (48.6) 24/145 (16.6) 4.76 [2.15; 10.54] <0.001
Type of bedding:
• medium warmth type 18/35 (57.1) 62/145 (42.7) 1.00 (reference)
• warm type 14/35 (40.0) 80/145 (55.2) 0.54 [0.25; 1.16] 0.114
• unknown type 1/35 (2.9) 3/145 (2.1) 1.03 [0.10; 10.50] 0.978
Bed sharing with parent(s) for the whole night # 0/35 (0) 20/145 (13.8) n/a 0.010
Sleeping in a room alone# 32/35 (91.4) 135/145 (93.1) n/a 0.268
INFANT FACTORS
Male sex 23/35 (65.7) 86/145 (59.3) 1.32 [0.61; 2.85] 0.487
Birth weight < 2500.0 g# 5/35 (14.3) 2/145 (1.4) n/a 0.003
Gestation ≤36 weeks# 3/35 (8.6) 2/145 (1.4) n/a 0.051
Any neonatal problem# 3/35 (8.6) 12/145 (8.3) n/a 0.589
Any congenital anomaly# 2/35 (5.7) 12/145 (8.3) n/a 0.463
Twins birth# 0/35 (0) 1/145 (0.7) n/a 0.810
Bottle feeding of the birth 7/35 (20.0) 7/144 (4.9) 4.89 [1.59; 15.05] 0.006
No dummy when sleeping 18/35 (51.4) 57/145 (39.3) 1.63 [0.78; 3.43] 0.194
MATERNAL FACTORS
Unplanned pregnancy 30/35 (85.7) 59/145 (40.7) 8.75 [3.21; 23.85] <0.001
Late (month 4–9) or no prenatal care 16/35 (45.7) 24/144 (16.7) 4.25 [1.91; 9.41] <0.001
Previous stillbirth# 2/35 (5.7) 18/144 (12.5) n/a 0.204
Previous interruption of pregnancy# 3/35 (8.6) 12/145 (8.3) n/a 0.589
Previous infant death# 1/35 (2.9) 4/144 (2.8) n/a 0.668
≥2 previous live births 26/35 (74.3) 74/144 (51.4) 2.73 [1.19; 6.24] 0.017
SOCIOECONOMIC FACTORS
Maternal age ≤20 y or ≥35 y 15/35 (42.9) 24/145 (16.6) 3.78 [1.69; 8.41] 0.001
Paternal age ≤20 y or ≥35 y 12/33 (36.4) 21/135 (15.6) 3.10 [1.33; 7.25] 0.009
Maternal education ≤12 y 25/35 (71.4) 43/145 (29.7) 5.93 [2.62; 13.40] <0.001
Paternal education ≤12 y 19/33 (57.6) 56/143 (39.2) 2.11 [0.98; 4.54] 0.057
Single mother*** 7/35 (20.0) 11/145 (7.6) 3.04 [1.09; 8.54] 0.034
No waged income 9/33 (27.3) 15/141 (10.6) 3.15 [1.24; 8.02] 0.016
Renting or living with parents 16/35 (45.7) 70/145 (48.3) 0.90 [0.43; 1.89] 0.785
Worse than fair housing conditions# 10/35 (28.6) 4/144 (2.8) n/a <0.001
Overcrowding**** # 32/35 (91.4) 110/144 (76.4) n/a 0.035
Worse than fair self-perceived economic family status 25/35 (71.4) 16/145 (11.0) 20.16 [8.21; 49.51] <0.001
PARENTAL LIFESTYLE FACTORS
Maternal smoking during pregnancy 10/35 (28.6) 8/145 (5.5) 6.85 [2.46; 19.05] <0.001
Paternal smoking during pregnancy 26/35 (74.3) 73/145 (50.3) 2.85 [1.25; 6.50] 0.013
Exposure to smoking during pregnancy (one or both of parents smoked) 28/35 (80.0) 73/145 (50.3) 3.94 [1.62; 9.61] 0.003
*Variables are for routine practices.
**Includes pallet, twisted plaid, pillow
***Includes mothers who were divorced, widowed, unmarried and living separately from the father of infant.
****Living >1 person per room.
#Fisher's exact test is carried out, hence odds ratio not applicable (n/a).
Over 30 factors were analyzed, and more than half of them were significant. Factors significant on univariate analysis were routine sleeping in a bassinet, routine use of substandard infant mattresses, such as pallet, twisted plaid or pillow for sleep, routine use of waterproof cloth over the mattress for sleep, routine wearing of ≥4 togs during sleep, routine wearing of cap during sleep, routine bed sharing with parent(s) for the whole night, birth weight <2500.0 g, gestation ≤36 weeks, bottle feeding off the birth, unplanned pregnancy, late (initial prenatal visit in month 4–9 of pregnancy) or no prenatal care, ≥2 previous live births, maternal age ≤20 or ≥35 years, paternal age ≤20 or ≥35 years, maternal education ≤12 years, single mother (those who were divorced, widowed, unmarried and living separately from the father of an infant), households without waged income, worse than fair housing conditions, worse than fair self-perceived economic family status, overcrowding (>1 person per room), maternal smoking during pregnancy, paternal smoking during pregnancy, and exposure to smoking during pregnancy when just one or both of the parents smoked.
Examples of factors that were not significant on univariate analysis included routine sleeping positions, routine sleeping in a full-sized bed (parental bed or sofa/couch), routinely used bedding type, routine use of pillow, routine sleeping in room alone, male sex, any neonatal problem, any congenital problem, twin birth, no dummy when sleeping, previous stillbirth, previous interruption of pregnancy, previous infant death, paternal education ≤12 years, and households renting or living with parents.
Multivariate results
A multivariate analysis was carried out to determine which factors were independently significant when controlled for other factors found to be important in the study. In the course of correlation analysis some correlates have been excluded from the multivariate analysis. So, all variables with an expected cell <5 have been excluded from the multivariate analysis. Odds ratios and p values for factors that remained significant or not significant after controlling for other significant on univariate analysis factors are presented in Tables 2 and 3.
Table 2 Multivariate analysis of significant factors in sleeping environment for risk of sudden infant syndrome
Variables* Univariate Multivariate
Sleeping+ All factors++
Odds ratio (95% CI) p Odds ratio (95% CI) p Odds ratio (95% CI) p
Sleeping in a bassinet 8.95 [3.80; 21.05] <0.001 6.54 [2.45; 17.42] <0.001 4.22 [1.16; 15.38] 0.029
Substandard infant mattress for sleep** 3.09 [1.38; 6.89] 0.006 3.36 [1.27; 8.84] 0.014 2.25 [0.69; 7.34] 0.180
Waterproof cloth over the mattress 2.35 [1.11; 4.98] 0.025 3.83 [1.43; 10.25] 0.007 2.87 [0.81; 10.21] 0.103
≥4 togs during sleep 8.02 [3.55; 18.16] <0.001 7.93 [3.16; 19.90] <0.001 8.49 [2.38; 30.32] 0.001
*Variables are for routine practices.
**Includes pallet, twisted plaid, pillow
+ Controlled for all sleeping environment factors listed in table.
++ Controlled for all factors that remained significant in the univariate analysis, including factors of sleeping environment.
Table 3 Multivariate analysis of significant factors in background characteristics for risk of sudden infant syndrome
Variables Univariate Multivariate
Background+ All factors++
Odds ratio (95% CI) p Odds ratio (95% CI) p Odds ratio (95% CI) p
Bottle feeding of the birth 4.89 [1.59; 15.05] 0.006 2.74 [0.45; 16.75] 0.274 2.43 [0.17; 33.74] 0.509
Unplanned pregnancy 8.75 [3.21; 23.85] <0.001 6.98 [2.19; 22.24] 0.001 5.22 [1.49; 18.18] 0.009
Late (month 4–9) or no prenatal care 4.25 [1.91; 9.41] <0.001 3.65 [1.24; 10.71] 0.019 2.49 [0.73; 8.51] 0.147
≥2 previous live births 2.73 [1.19; 6.24] 0.017 2.01 [0.69; 5.86] 0.199 3.90 [1.00; 15.10] 0.048
Maternal age ≤20 or ≥35 y 3.78 [1.69; 8.41] 0.001 1.98 [0.67; 5.89] 0.216 1.18 [0.34; 4.04] 0.792
Maternal education ≤12 y 5.93 [2.62; 13.40] <0.001 4.11 [1.48; 11.41] 0.007 4.48 [1.34; 14.94] 0.015
Single mother 3.04 [1.09; 8.54] 0.034 1.86 [0.36; 9.61] 0.457 1.87 [0.22; 16.06] 0.566
No waged income 3.15 [1.24; 8.02] 0.016 1.01 [0.26; 3.86] 0.993 1.04 [0.21; 5.10] 0.965
Exposure to smoking during pregnancy (one or both of parents smoked) 3.94 [1.62; 9.61] 0.003 2.76 [0.87; 8.71] 0.083 1.87 [0.51; 6.94] 0.348
+ Controlled for all background factors listed in table.
++ Controlled for all factors that remained significant in the univariate analysis, including background factors.
Table 2 shows how the significance of the variables associated with the sleeping environment changes when they are put in the multivariate model with each other and how it changes further when we control for other significant risk factors. Table 3 shows how the significance of the background variables associated with the infant and maternal medical history, parental socioeconomic status and lifestyle changes when they are put in the multivariate model with each other and how it changes further when we control for other significant risk factors. The risk associated with routine sleeping in a bassinet, routinely practiced heavy wrapping (≥4 togs) of an infant during sleep, unplanned pregnancy, 2 or more previous live births and maternal education ≤12 years remained significant when we controlled for all factors. The risks associated with a substandard infant mattress and a waterproof cloth over the mattress for sleep were significant among the factors associated with the sleeping environment but just failed to reach significance when we controlled for all risk factors. The risk associated with a late (month 4–9) or absent prenatal care was significant among the background factors but failed to reach significance when we controlled for all risk factors. The risks associated with a bottle feeding of the birth, maternal age ≤20 y or ≥35 y, single mother, households having no waged income and exposure of a fetus to smoking during pregnancy failed to reach significance both when we controlled for background and for all risk factors.
Later on, models were constructed with the backward stepwise procedure for variables significant at the 5% level. The multivariate model with the best prognostic power of multivariate logistic regression analysis including all effect modifiers that were significant in the univariate analysis and remained significant in the multivariate analysis was produced on the 8th step (Table 4). Therefore, we have established that the variables, such as unplanned pregnancy, routine use of a waterproof cloth over the mattress for sleep, routinely practiced heavy wrapping (≥4 togs) of an infant during sleep, maternal education ≤12 years, 2 or more previous live births and routine sleeping in a bassinet being together significantly predict SIDS victim.
Table 4 Associations of variables in the multivariate model with the best prognostic power (on step 8th)
Variables Odds ratio (95% CI) p
Unplanned pregnancy 6.22 [1.93; 20.07] 0.002
Waterproof cloth over the mattress* 2.65 [0.87; 10.35] 0.101
≥4 togs during sleep* 9.51 [2.91; 31.05] <0.001
Maternal education ≤12 y 5.97 [1.98; 17.96] 0.001
≥2 previous live births 5.38 [1.52; 19.06] 0.009
Sleeping in a bassinet* 5.14 [1.61; 16.36] 0.006
*Variables are for routine practices.
Discussion
Till now no observational analytic study concerning SIDS had been performed in Lithuania and data about SIDS were fragmental. Our study has disclosed some specific risk factors significantly associated with and predicting increased risk of SIDS relevant in Lithuania. SIDS group of 35 cases was small but sufficient to yield moderate effects with a power of 0.8 for general tests. Low mortality rate from SIDS in Lithuania resulted in a retrospective case-control study. The response rate of both SIDS parents and control parents was high. Matching SIDS victims and controls for the date of birth and region of birth further reduces confounders. For both SIDS cases and controls there was a time lag between the period of the questions related to and the time of the actual questionnaire. We have considered the possibility that some of the associations are the result of a recall bias. However, we believe the effect of a recall bias was minimal as in other retrospective and prospective studies, recall bias has been found not to influence the results. We have considered the possibility of an informational bias too, because of different ways of information collection. However, the effect of informational bias was minimal as during home visits only standardized questions for SIDS cases were asked. Home visits for SIDS cases were made because of low response of cases during other case-control studies regarding deaths in Lithuania.
Seasonality, with an increased incidence during winter and a decreased incidence during summer, has been considered a distinctive feature of SIDS. We have found a higher SIDS incidence in warm season. Before a definitive answer can be given on the role of the seasons, an extensive analysis should be performed on national data, taking into account the seasonality of births and age effects.
Importantly, this investigation has disclosed that prone sleeping position is not a risk factor for SIDS in Lithuania. The average mortality rate from SIDS in Lithuania is 0.3 per 1000 live births and is low, if compared to international mortality rates. Possibly, it is related to rare prone sleeping, prevalence rate of which is not higher than 3–4% [6]. The countries that have achieved prone prevalence rates of 3–10%, following the introduction of risks reducing campaigns, have SIDS mortality rates as low as 0.4–0.5 per 1000 live births [4].
The side sleeping position as a risk factor for SIDS in Lithuania has not been confirmed, either. The side (73%) and back (17%) positions are commonly recommended sleep positions at discharge from maternity unit in Lithuania. Epidemiological studies from England and New Zealand have shown that side sleeping has a slightly higher risk of SIDS than the supine position, though not as great as prone sleeping [7,8]. The higher risk for SIDS among infants placed on their sides may be related to a relative instability of this position. Although infants placed on their sides usually roll to their backs, the risk of rolling to the prone position from the side is significantly greater than rolling to the prone position from the back. Since sleeping on the back is the safest, parents in Lithuania should be prompted to use this position wherever possible.
Other results from the sleep surface analysis were unexpected for us, too. The risk associated with routine sleeping in a bassinet, a substandard infant mattress for sleep and a waterproof cloth over the mattress during sleep were significant among the factors associated with the sleeping environment. Routine sleeping in a bassinet and waterproof cloth over the mattress have entered in the multivariate model with the best prognostic power. Sleeping in bassinets, which are designed mainly for carriage, unlike cots, which are designed for sleep and meet safety standards for infants, may, at least in theory, carry a risk of accidental entrapment and suffocation. As we have found only a low correlation between placing an infant to sleep in a bassinet and households having no waged income, this practice is not a marker for a low economic status, but a stereotype of a routine infant sleeping environment in Lithuania. A waterproof cloth over the mattress during sleep as well as other loose bedding may carry a risk of accidental entrapment and suffocation, too. Epidemiological studies worldwide identified soft surfaces, such as pillows, quilts, comforters, sheepskins and porous mattresses as a significant risk factor, particularly when placed under a sleeping infant [9,10]. However, the risk associated with sleeping on substandard infant mattresses, such as a pallet or a twisted plaid is somewhat disturbing and needs further investigation. So, the findings from sleeping environment analysis are somewhat puzzling, but illustrate the fact of culturally diverse infant care practices.
Some studies investigating SIDS have reported the risk associated with the excessive amount of clothing or bedclothes and high-temperature environments during sleep, particularly for infants lying prone [11]. In our study a routinely practiced heavy wrapping (clothing, caps and socks) of an infant during sleep emerged from a multivariate analysis as one of the most important independent SIDS risk factors, though most infants routinely slept on their sides or supine. The increased SIDS risk associated with overheating is particularly evident when infants sleep prone but is less clear when they sleep supine [11]. As most infants in the present study slept on their sides or supine, adverse effects of heavy wrapping would be less likely. So our findings suggest that interaction between the heavy wrapping of an infant, especially when the head is covered with a cap, and sleeping in a bassinet may exist and relate with a bigger thermal stress.
Some scientific studies have demonstrated that bed sharing can alter sleep patterns of mother and baby [12,13]. These studies have led to speculations that bed sharing may also reduce the risk of SIDS. While bed sharing may have certain benefits such as encouraging breastfeeding there are not studies demonstrating that bed sharing reduces SIDS. Some studies actually suggest that bed sharing under certain conditions such as smoking parents may increase the risk of SIDS [14]. Our study can not confirm or reject these speculations as bed sharing was common only for the controls.
The contribution of artificial feeding to SIDS is not clear and may vary in different communities. On the basis of our results bottle feeding from the birth is not a risk factor for SIDS when controlled for other significant factors on multivariate analysis. However, there are other good reasons to continue promoting breastfeeding.
The higher incidence of cases with low birth weight and preterm birth confirms findings of other studies [15].
Unplanned pregnancy and 2 or more previous live births are significant potentially modifiable risk factors for SIDS found in our study. Although these factors do not point to a specific etiology, they suggest that interventions targeted towards family planning methods may theoretically reduce SIDS incidence in Lithuania.
The data from univariate analysis related with socioeconomic household status, such as parental age at the birth of an infant, parental education, waged income, housing conditions, number of persons living per room and self-perceived parental economic situation have demonstrated it to be significantly lower in SIDS group than in control. Further, the multivariate analysis showed maternal education ≤12 years to be one of the very significant independent SIDS risk factors in Lithuania. Mothers with lower education might care differently for their babies than mothers having higher education, and they may possibly have many worries, other than those about their child. A link between low socioeconomic status and SIDS has been noted in the literature [16].
Maternal smoking during pregnancy is a major, potentially modifiable risk factor found in many other studies [14,17,18]. Epidemiologically it is difficult to distinguish the effect of active maternal smoking during pregnancy from involuntary tobacco smoking so we have analyzed exposure to smoking according to smoking of partners. Even though the exposure to smoking during pregnancy was significant in the univariate model, we found no significant difference in the multivariate model.
The variables found to be significant in case-control study depend on what is included in a multivariate model. Further more extensive and detailed epidemiological research of SIDS is needed in Lithuania in order to understand the complex relationships between these variables and other factors that affect infant health and serve to heighten the risk of SIDS.
Conclusion
The results of this first population-based case-control study of the SIDS in Lithuania have shed some light on the epidemiology of the syndrome in Lithuania. The circumstance that some of these factors may be modifiable has important implications in terms of social policy and health education. Although the mortality of SIDS in Lithuania is not high, it might be lowered moreover by public informing about SIDS and related risk factors. Special attention must be paid to mothers with low education on potentially modifiable risk factors such as routine heavy wrapping of an infant during sleep, routine sleeping in a bassinet and unplanned pregnancy.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
The study was conceived and designed by Vilija Bubnaitienė and Ramunė Kalėdienė with assistance from Rimantas Kėvalas. The study was conducted under the supervision of Ramunė Kalėdienė. Statistical analysis was conducted by Vilija Bubnaitienė with assistance from Ramunė Kalėdienė and Irena Nedzelskienė. Interpretation of results was conducted by Vilija Bubnaitienė with assistance from Ramunė Kalėdienė. The manuscript was prepared by Vilija Bubnaitienė and edited by Ramunė Kalėdienė and Rimantas Kėvalas. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
We would like to thank all the parents who participated in this study, and medical statistician Irena Nedzelskienė for support in statistical analysis.
==== Refs
Burnett LB Adler J Sudden Infant Death Syndrome eMedicine J 2001 2
Kaarene Fitzgerald AC SIDS Global Strategy Task Force: international SIDS statistics (updated 24th August 2000)
AAP Task Force on Infant Sleep position and Sudden Infant Death Syndrome Changing concepts of sudden infant death syndrome: implications for infant sleeping environment and sleep position Pediatrics 2000 105 650 656 10699127 10.1542/peds.105.3.650
Beal SM Rognum TO Sleeping position and SIDS: past, present and future Sudden Infant Death Syndrome: New Trends in the Nineties 1995 Scandinavian University Press, Oslo, Norway 147 151
Willinger M James LS Catz C Defining the sudden infant death syndrome (SIDS): deliberation of an expert panel convened by the National Institute of Child Health and Human Development Pediatric Pathology 1991 11 677 684 1745639
Nelson EA Serra A Cowan S Maternity advice survey: sleeping position in Eastern Europe Arch Dis Child 2000 83 304 306 10999862 10.1136/adc.83.4.304
Fleming PJ Blair PS Bacon C Environment of infants during sleep and risk of the sudden infant death syndrome: results of 1993–5 case-control study for confidential inquiry into stillbirths and deaths in infancy BMJ 1996 313 191 195 8696193
Scragg RK Mitchell EA Side-sleeping position and bedsharing in the sudden infant death syndrome Ann Med 1998 30 345 349 9783832
Mitchell EA Thompson JMD Ford RPK Sheepskin bedding and the sudden infant death syndrome J Pediatr 1998 133 701 704 9821434 10.1016/S0022-3476(98)70116-7
Mitchell EA Scragg L Clements M Soft cot mattresses and the sudden infant death syndrome N Z Med J 1996 109 206 207 8668299
Fleming PJ Levine MR Azaz Y Wigfield R The development of thermoregulation and interactions with the control of respiration in infants: possible relationship to sudden infant death Acta Pediatr Scand 1993 57 59
Mosco SS Richard CA McKenna JJ Infant arousals during mother-infant bed sharing: implications for infant sleep and sudden infant death syndrome research Pediatrics 1997 100 841 849 9346985 10.1542/peds.100.5.841
Young J Sawczenko A Fleming PJ Night-time behavior between low SIDS risk infants and their mothers: a longitudinal study of room-sharing and bed-sharing Pediatr Pulmonology 1996 22 429
Scragg RKR Mitchell EA Bedsharing, smoking and alcohol in the sudden infant death syndrome BMJ 1993 307 1312 1318 8257885
Hoffman HJ Hillman LS Epidemiology of the sudden infant death syndrome: Maternal, neonatal, and postneonatal risk factors Clin Perinatol 1992 19 717 737 1464187
Leach CE Blair PS Fleming PJ Epidemiology of SIDS and Explained Sudden Infant Deaths Pediatrics 1999 104 43 55 10390258 10.1542/peds.104.4.e43
Blair P Fleming PJ Benslay D Smoking and the sudden infant death syndrome: results from 1993–5 case-control study for confidential inquiry into stillbirths and deaths in infancy BMJ 1996 313 195 198 8696194
Mitchell EA Rognum TO Smoking: the next major and modifiable risk factor Sudden infant death syndrome New Trends for the nineties 1995 Oslo, Norway: Scandinavian University Press 114 118
|
16283946
|
PMC1308821
|
CC BY
|
2021-01-04 16:31:07
|
no
|
BMC Pediatr. 2005 Nov 13; 5:41
|
utf-8
|
BMC Pediatr
| 2,005 |
10.1186/1471-2431-5-41
|
oa_comm
|
==== Front
BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-1181628394010.1186/1471-2458-5-118Research ArticlePrevalence of Helicobacter pylori infection and associated factors among adults in Southern Brazil: a population-based cross-sectional study Santos Ina S [email protected] Jose [email protected] Ari S [email protected] Neiva CJ [email protected] Camila S [email protected] Marta Colvara [email protected] Ricardo D [email protected] Department of Social Medicine, Faculty of Medicine, Federal University of Pelotas, Pelotas, Brazil2 Laboratory of Stable Isotopes applied to Medicine and Biology, School of Pharmacy and Biochemistry, University of Buenos Aires, Buenos Aires, Argentina3 Department of Analytic and Inorganic Chemistry, Instituto f Chemistry and Geosciences, Federal University of Pelotas, Pelotas, Brazil2005 10 11 2005 5 118 118 3 6 2005 10 11 2005 Copyright © 2005 Santos et al; licensee BioMed Central Ltd.2005Santos et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Helicobacter pylori (Hp) infection is associated with several upper gastrointestinal disorders. Local data on the epidemiology of the infection are scarce in Brazil. The purpose of this study is to measure the prevalence rate and to explore the associated factors among the adult population living in Pelotas, a southern Brazilin city.
Methods
This was a population-based cross-sectional study. Through a multi-stage sampling method all individuals 20 years and over living at the selected households at the urban area of the city were interviewed regarding past and current socio-economic indicators; demographic characteristics; nutritional and behavioural habits; and history of upper gastrointestinal symptoms.Hp infection was ascertained through the 13C-UBT. Due to the high prevalence, data were analysed through robust Poisson regression. All analyses took into account the family clustering of the data.
Results
Among 563 eligible individuals, 363 agreed to perform the 13C-UBT (refusal rate of 35.5%). Refusals were associated with female sex, consumption of mate drinking, and presence of upper gastrointestinal symptoms. The prevalence rate of H. pylori infection was 63.4% (95%CI 59.3%–69.3%). In crude analyses, prevalence was associated with increasing age, non-white skin colour, lower current family income, lower education level, higher size of the family, low socio-economic conditions in childhood, higher number of siblings and attendance to day-care centres in childhood, and presence of dyspeptic symptoms. In adjusted analysis the level of education of the father was inversely associated with the infection, whereas number of siblings and attendance to day-care centre in childhood were directly associated with it. Non-white skin colour remained significantly associated with increased prevalence even after allowing for past and current socio-economic characteristics, age and sex. Compared to non-symptomatic individuals, those reporting dyspeptic symptoms presented a higher prevalence of the infection even after allowing for current and past socio-economic conditions, ethnicity, age, and sex.
Conclusion
Hp infection is as common among adults in southern Brazil as it is in other developing countries. Socio-economic conditions in childhood besides ethnicity and presence of dyspeptic symptoms were the factors significantly associated with the infection.
==== Body
Background
Helicobacter pylori (H. pylori) infection is chronic and common throughout the world, with a higher prevalence in developing than in developed countries [1]. H. pylori infection causes gastritis and is the most important risk factor for peptic ulcer disease (gastric and duodenal)[2,3]. It also contributes to the onset of gastric cancer and primary gastric B-cell lymphoma [2,4] and more recently has been investigated as a risk factor for ischemic heart disease[5] and intrauterine growth restriction[6,7].
In Brazil, in 1998, among the malignancies, gastric cancer was second only to lung cancer as the cause of mortality in men[8]. Despite of this, local epidemiological data on H. pylori infection are scarce. Through a non-restrictive search at Medline and SciElo databases using "Helicobacter pylori" and "prevalence or seropositivity" and "Brazil or Brazilian" as title/abstract descriptor terms, only seven studies primarily planned to measure prevalence of the infection among adult non-patient individuals were found [9-15]. Five of these studies had been conducted among selective populations (abattoir workers[9], Japanese Brazilians and Japanese residents in Brazil[11,12], blood donors[14], and native populations from Brazilian Western Amazon)[15]. Overall prevalence rate in those studies varied from 48% among Japanese Brazilians living in four different cities in Brazil[12] to 84.7% among the adult residents of a rural area of the state of Mato Grosso[10].
This paper reports the prevalence rate of the H. pylori infection and the factors that showed to be associated with the infection in a study conducted in Pelotas, a city of 320,000 inhabitants, located in the state of Rio Grande do Sul (southern Brazil).
Methods
This was a population-based cross-sectional study. A multi-stage sampling method, taking the structure of the population as a basis, was used for sample selection. Fifty eight of the 404 census tracts of the city were randomly chosen and, subsequently, a random sample of five households per census tract was selected to the study. All adults (20 years and over) living in the selected households were invited to participate. Unoccupied houses were excluded and the house next-door was selected. In the case of temporary absence of one or more adults from the household (students, workers) the interviewer returned later, or during the next day, to perform the interview. After three unsuccessful attempts of contact, at different days, the adult/household was considered a loss. Refusals were also considered losses.
Subjects were interviewed at home by trained field workers using a standard questionnaire. Information on current (family monthly income in minimum wages, educational level, marital status, and size of the family) and past socio-economic characteristics (history of having lived only in urban or in urban and rural areas, level of formal education of the parents of the interviewee, number of siblings, and attendance to day-care centres in childhood) was collected.
Age was gathered in complete years. Gender and ethnicity (as defined according to skin colour) were observed by the interviewer. History of upper abdominal complaints in the last year: dyspeptic symptoms (pain or discomfort), substernal burning pain, decreased appetite, abdominal fullness, abdominal bloating, early satiety, and vomiting were investigated. History of peptic ulcer disease and gastric cancer in first degree relatives was explored.
Frequency of consumption of raw vegetables, coffee, alcoholic beverages and mate (Ilex paraguayensis) were collected. Smoking history was also provided by the interview. Changes in nutritional habits, alcoholic beverages ingestion and in smoking pattern due to the presence of upper abdominal symptoms were also investigated.
H. pylori infection was assessed by the isotope technique, the 13C-Urea Breath Test (13C-UBT) after at least six hours fastening. Each dose consisted into about 50 gram portion of 13C-urea (Cambridge Isotope Laboratories Inc., Massachusetts, USA). The samples were measured in a mass spectrometer coupled to a gas chromatographer (FinniganMAT GmbH, ThermoQuest Corp., Bremen, Germany). Subjects with an excess δ13CO2 value of > 3.5 per ml were defined as H. pylori positive[16].
Due to the high prevalence of the H. pylori infection, prevalence rates, prevalence ratios, and the respective 95% confidence intervals were calculated through robust Poisson regression[17], taking the family clustering of the data into account. The Stata 8.0 package was used for these analyses. A hierarchical model of analysis[18] was applied according to a conceptual framework established a priori, through which caudal variables were adjusted for cranial ones. At the first most cranial level were the socio-economic characteristics in childhood. At the second level, the current socio-economic variables, followed at the third level by the demographic factors. In the fourth and fifth levels, lifestyle characteristics and complaints of upper gastrointestinal symptoms, respectively, were entered to the model. Variables significant at their level were kept in the model even whether the inclusion of caudal variables turned their association with the outcome statistically non significant. At every level, a backward selection of variables was performed. Those presenting a p value ≤ 0.20 were kept in the model for adjustment purpose. Only variables with p value < 0.05 were considered significantly associated with H. pylori infection.
The study protocol was cleared by the Ethical Committee of the Federal University of Pelotas. A written informed consent was obtained from participants before enrolment in the study.
Results
Test compliers versus non-compliers
A total of 277 households were visited. In all, 563 adults answered the questionnaire and of them 363 accepted to have the 13C-UBT performed, a refusal rate of 35.5%. Tables 1, 2, 3, 4 describe compliers and non-compliers according to the independent variables investigated. Compliance to the test was significantly (p < 0.05) associated with sex (higher among the women), mate intake (higher among the consumers), and presence of most of the gastrointestinal symptoms explored (higher among the symptomatic individuals). The statistical significance of the association between compliance and current family income was borderline (p = 0.09).
Table 1 Demographic and socio-economic characteristics of non-complier and complier subjects, and prevalence of H. pylori infection
Characteristics Test non-compliers (n = 200) n (%) Test compliers (n = 363) n (%) p1 Prevalence2 n (%) p3
Age (years) 0.6 0.024
20–29 53 (26.5) 84 (23.4) 40 (47.6)
30–39 35 (17.5) 78 (21.7) 56 (71.8)
40–49 42 (21.0) 80 (22.3) 57 (71.3)
≥ 50 70 (35.0) 117 (32.6) 78 (66.7)
Sex 0.009 0.1
Male 95 (47.5) 129 (35.9) 76 (58.9)
Female 105 (52.5) 230 (64.1) 155 (67.4)
Ethnicity 0.6 <0.001
White 151 (75.5) 279 (77.7) 165 (59.1)
Non white 49 (24.5) 80 (22.3) 66 (82.5)
Family income (MW) 0.09 <0.0014
≤ 1.00 27 (14.1) 37 (10,5) 31 (83.8)
1.01–3.00 66 (34.4) 138 (39.0) 92 (66.7)
3.01–5.00 42 (21.9) 99 (28.0) 71 (71.7)
≥ 5.01 57 (29.7) 80 (22.6) 35 (43.8)
Education 0.5 0.0014
≤ 4 39 (21.5) 69 (19,2) 49 (71.0)
5–7 41 (22.7) 84 (23.4) 60 (71.4)
8–10 32 (17.7) 72 (20.1) 48 (66.7)
≥ 11 69 (38.1) 109 (30.4) 53 (48.6)
Marital status 0.4 0.8
With partner 127 (63.5) 241 (67.1) 156 (64.7)
Without partner 73 (36.5) 118 (32.9) 75 (63.6)
Size of present family 0.7 0.03
1–3 92 (46.0) 171 (47.6) 100 (58.5)
≥ 4 108 (54.0) 188 (52.4) 131 (69.7)
1p of difference between test-compliers and test non-compliers according to the independent variables
2 prevalence among test-compliers
3p of the association between the independent variable and the prevalence of H. pylori infection
4test for linear trend
Table 2 Past socio-economic among non-complier and complier subjects, and prevalence of H. pylori infection.
Characteristics Test non-compliers (n = 200) n (%) Test compliers (n = 363) n (%) p1 Prevalence2 n (%) p3
Place of living 0.1 0.2
Urban only 133 (66.5) 217 (60,4) 135 (62.2)
Urban and rural 66 (33.5) 142 (39.6) 96 (67.6)
Mother education 0.6 0.0014
≤ 4 108 (63.9) 203 (65.1) 143 (70.4)
5–7 29 (17.2) 61 (19.6) 38 (62.3)
8–10 13 (7.7) 16 (5.1) 8 (50.0)
≥ 11 19 (11.2) 32 (10.3) 12 (37.5)
Father education (y) 0.2 0.0014
≤ 4 101 (63.9) 181 (50.4) 127 (70.2)
5–7 21 (13.3) 59 (16,4) 39 (66.1)
8–10 12 (7.6) 24 (6.7) 11 (45.8)
≥ 11 24 (15.2) 30 (8.4) 9 (30.0)
N° siblings 0.9 0.0014
0–2 60 (30.2) 113 (31.5) 55 (48.7)
3–4 49 (24.6) 77 (21.4) 54 (70.1)
5–6 35 (17.6) 67 (18.7) 43 (64.2)
≥ 7 55 (27.6) 102 (28.4) 79 (77.5)
Day-care center childhood 0.2 <0.001
Yes 11 (5,5) 11 (3.1) 10 (90.9)
No 188 (94.5) 348 (96.9) 221 (63.5)
1p of difference between test-compliers and test non-compliers according to the independent variables
2 prevalence among test-compliers
3p of the association between the independent variable and the prevalence of H. pylori infection
4test for linear trend
Table 3 Current lifestyle among non-complier and complier subjects, and prevalence of H. pylori infection.
Characteristics Test non-compliers (n = 200) n (%) Test compliers (n = 363) n (%) p1 Prevalence2 n (%) p3
Smoking 0.6 0.34
Never smoked 99 (49.7) 164 (45.7) 99 (60.4)
Ex-smoker 44 (22.1) 88 (24.5) 63 (71.6)
Current smoker 56 (28.1) 107 (29.8) 69 (64.5)
Alcohol intake 0.9 0.07
No 101 (50.8) 186 (51.8) 128 (68.8)
Yes 98 (49.2) 173 (48.2) 103 (59.5)
Coffee intake (d/wk) 0.4 0.024
0 27 (13.6) 38 (10.6) 18 (47.4)
1–6 21 (10.6) 43 (12.0) 25 (58.1)
7 151 (75.9) 278 (77.4) 188 (67.6)
Mate intake (d/wk) 0.03 0.34
0 70 (35.2) 81 (22.6) 58 (71.6)
1–6 39 (19.6) 101 (28.2) 59 (58.4)
7 90 (45.2) 176 (49.2) 113 (64.2)
Raw vegetables (d/wk) 0.2 0.64
0 34 (17.2) 51 (14.4) 34 (66.7)
1–6 113 (57.1) 188 (53.1) 116 (61.7)
7 51 (25.8) 115 (32.5) 78 (67.8)
1p of difference between test-compliers and test non-compliers according to the independent variables
2 prevalence among test-compliers
3p of the association between the independent variable and the prevalence of H. pylori infection
4test for linear trend
Table 4 Upper gastrointestinal symptoms among non-complier and complier subjects, and prevalence of H. pylori infection.
Characteristics Test non-compliers (n = 200) n (%) Test Compliers (n = 363) n (%) p1 Prevalence2 n (%) p3
Dyspeptic symptoms 0.05 0.01
No 172 (86.4) 286 (79.7) 176 (61.5)
Yes 27 (13.6) 73 (20.3) 55 (75.3)
Substernal burning 0.01 0.09
No 108 (54.5) 155 (43.2) 92 (59.4)
Yes 90 (45.5) 204 (56.8) 139 (68.1)
Decreased appetite 0.001 0.2
No 164 (82.4) 246 (68.9) 154 (62.6)
Yes 35 (17.6) 111 (31.1) 77 (69.4)
Abdominal fullness 0.001 0.3
No 188 (94.5) 307 (85.5) 195 (63.5)
Yes 11 (5.5) 52 (14.5) 36 (69.2)
Early satiety 0.1 0.8
No 192 (96.5) 335 (93.3) 216 (64.5)
Yes 7 (3.5) 24 (6.7) 15 (62.5)
Vomiting 0.06 0.6
No 194 (97.5) 337 (93.9) 218 (64.7)
Yes 5 (2.5) 22 (6.1) 13 (59.1)
Abdominal bloating 0.03 0.5
No 172 (86.9) 285 (79.4) 181 (63.5)
Yes 26 (13.1) 74 (20.6) 50 (67.6)
Family history peptic disease 1° degree relatives 0.7 0.9
No 23 (43.4) 57 (40.7) 36 (63.2)
Yes 30 (56.6) 83 (59.3) 53 (63.9)
Family history gastric cancer 1°degree relatives 0.6 0.9
No 25 (83.3) 57 (68.7) 39 (68.4)
Yes 5 (16.7) 26 (31.3) 18 (69.2)
1p of difference between test-compliers and test non-compliers according to the independent variables
2 prevalence among test-compliers
3p of the association between the independent variable and the prevalence of H. pylori infection
At the whole sample (n = 563) no statistical association between sex and mate drinking was detected: 73.1% and 72.8%, respectively, of the men and women were regular mate drinkers (at least once a week). The association between mate drinking and dyspeptic symptoms was in the limit of the statistical significance both in crude (p = 0.05) and in adjusted analysis (p = 0.06): 19.9% of the mate drinkers reported dyspeptic symptoms against 12.6% of the non-drinkers. The same was observed with the association between sex and dyspepsia (p = 0.9): the prevalence of dyspeptic symptoms was slightly higher among women (20.3%) than among men (14.3%). Stratified analysis according to the compliance status of the interviewees did not change these findings, except by the higher p-values.
Description of the compliers
On Table 1, among the compliers, almost two thirds were women (64.1%) and lived with a partner (67.1%). More than 50% shared household with four or more relatives and 49.5% were from families earning no more than three Brazilian minimum wages per month. Almost one third of the compliers (30.4%) had more than ten years of formal education. Regarding past socio-economic characteristics (Table 2), 39.6% were born or spent part of their lives at the rural area. The level of education of the mother and the father for 65.1% and 50.4%, respectively, of the compliers was of four years or less. More than one fourth of the individuals (28.4%) belonged to families with seven or more siblings. A small proportion of the participants (3.1%) attended to day care centres in childhood.
Table 3 shows that 29.8% were smokers. Consumption of alcoholic beverages was reported by 48.2% of the interviewees. Daily coffee drinking was reported by 77.4% of the participants, while 49.2% informed to be daily mate drinkers. Consumption of raw vegetables at least once a week was reported by 85.6% of the individuals.
On Table 4, prevalence of dyspeptic symptoms (upper abdominal pain or discomfort in the last year) was of 20.3%. Other symptoms like substernal burning pain, decreased appetite, abdominal bloating, abdominal fullness, early satiety, and vomiting were complained in the last year by, respectively, 56.8%, 31.1%, 20.6%, 14.5%, 6.7%, and 6.1% of the interviewees. One hundred fifty seven individuals had sought medical care due to abdominal symptoms. Gastritis, peptic ulcer disease and gastro-oesophageal reflux were reported, respectively, by 38.9%, 9.3% and 7.4% of the 54 participants who were aware, at the moment of the interview, of the medical diagnosis to their symptoms.
Crude prevalence rates and adjusted prevalence ratios for H. pylori infection
Only four of the 363 individuals who had the 13C-UBT had undetermined test result. Overall prevalence of H. pylori infection among the remaining 359 subjects was of 64.3% (95%CI 59.3%–69.3%). In crude analysis, prevalence of H. pylori infection was statistically associated with age, ethnicity, family monthly income, level of education, size of the actual family, and past socio-economic indicators (Tables 1 and 2). The prevalence increased with age although remaining relatively constant after the twenties: from 47.6% among individuals 20–29 years old to more than 70% among the 30–49 year old. Prevalence among non-white individuals was 40% higher than the observed among whites (respectively, 82.5% and 59.1%; p < 0.001). Infection rate among the poorest (83.8%) was almost twice as high as among those from families earning more than five minimum wages a month (43.8%). Association with educational level was reversed: the higher the education level the lower the prevalence of H. pylori infection (p = 0.001). Comparatively to individuals from smaller families (1–3 members), those living with four or more relatives had a prevalence rate of infection 19% higher (respectively, 58.5% and 69.7%). As with the participant level of education, association with both maternal and paternal education was reversed. Prevalence was 87% higher among those whose mother presented an educational level ≤ 4 years, comparatively to those whose mothers had ≥ 11 years of education (p = 0.001). The association with the paternal schoolarity was even stronger, with 2.3 fold increase in prevalence rate among individuals whose father had low schoolarity, comparatively to those with eleven years or more of formal education. A linear trend was observed between number of siblings in childhood and prevalence of the infection with higher prevalence according to the increase in the number of siblings. Prevalence was 43% higher among those who had attended day-care centres in childhood comparatively to those who did not (respectively, 90.9% and 63.5%).
Among lifestyle characteristics, only coffee intake was significantly associated with the infection. A linear trend was observed between weekly frequency of coffee consumption and the H. pylori infection rate (Table 3). Daily drinkers presented a prevalence rate of 67.6%, whereas among the non-drinkers the prevalence was 47.4%. The statistical significance of the association with alcohol consumption was borderline (p = 0.07), with non-consumers presenting a higher prevalence of infection than consumers (respectively, 68.8% and 59.5%).
On Table 4, only the presence of dyspeptic symptoms was significantly associated with the occurrence of H. pylori infection, with symptomatic individuals having a prevalence 22% higher than the observed among non-symptomatic ones (p = 0.01). The statistical significance of the association between substernal burning and the infection was borderline (p = 0.09): symptomatic individuals had a prevalence rate 15% higher than those not reporting this symptom. No association between family history of peptic ulcer disease or gastric cancer and H. pylori infection was detected.
Table 5 presents crude and adjusted prevalence ratios of the independent variables significantly associated with the outcome. Level of education of the father, number of siblings and attendance to day-care centres in childhood remained statistically significant after adjusting to all other past socio-economic indicators. Father education was protective against the occurrence of the infection. Subjects whose father had eleven years or more of formal education presented a 53% decrease in the probability of infection when compared to those whose father had ≤ 4 years of schoolarity. The occurrence of infection increased linearly with the number of siblings: individuals from larger families in childhood (≥ 7 siblings) presented a probability 55% higher of being H. pylori positive than those from families with no more than two children. Among those who attended day-care centres in childhood, the probability of the infection was 53% higher than among their counterparts.
Table 5 Crude and adjusted prevalence ratios of independent variables for H. pylori infection. (n = 359)
Characteristics Crude PR 95% CI Adjusted PR 95% CI p
Father education) 0.011
≤ 4 1.00 1.00
5–7 0.94 0.77 – 1.16 0.98 0.80 – 1.20
8–10 0.65 0.42 – 1.02 0.69 0.46 – 1.06
≥ 11 0.42 0.24 – 0.74 0.47 0.28 – 0.79
N° siblings 0.0011
0–2 1.00 1.00
3–4 1.44 1.13 – 1.83 1.44 1.11 – 1.86
5–6 1.32 1.02 – 1.71 1.27 0.96 – 1.68
≥ 7 1.59 1.28 – 1.98 1.55 1.22 – 1.98
Day-care centre in childhood 0.001
Yes 1.01 1.01 – 1.01 1.53 1.18 – 1.97
No 1.00 1.00 1.00
Current family income (MW) 0.071
≤ 1.00 1.00 1.00
1.01–3.00 0.80 0.66 – 0.96 0.86 0.69 – 1.06
3.01–5.00 0.86 0.71 – 1.03 0.97 0.79 – 1.19
≥ 5.01 0.52 0.39 – 0.69 0.68 0.49 – 0.93
Ethnicity <0.001
White 1.00 1.00
Non white 1.39 1.21 – 1.61 1.32 1.14 – 1.54
Dyspeptic symptoms 0.002
No 1.00 1.00
Yes 1.22 1.04 – 1.44 1.28 1.10 – 1.50
1test for linear trend
The present socio-economic variables significantly associated with H. pylori infection in crude analysis (current family income, education and size of the present family) lost their statistical significance after controlling for past socio-economic indicators. Family monthly income presented a borderline statistical significance (p = 0.07). Individuals from families earning five or more minimum wages presented a probability of infection 32% lower than those from the poorest families. However, the confidence limits of the adjusted prevalence ratios of all the other categories of the variable included the unit.
Among the demographic variables, after adjusting for past and present socio-economic factors, only ethnicity remained significantly associated with the outcome. The adjusted prevalence ratio for non-whites was 32% higher than that of the whites. Age which was linearly and significantly associated with the infection in crude analysis lost its significance after adjusting for confounders (p = 0.1). Together with sex (p = 0.1) the variable age was kept in the model for adjustment of the subsequent variables.
None of the lifestyle characteristics were significantly associated with H. pylori infection in adjusted analysis. Among the variables of this group, only weekly frequency of coffee intake (p = 0.1) was kept in the model for controlling the upper gastrointestinal symptoms. Individuals complaining of dyspeptic symptoms in the last year had an adjusted prevalence ratio 28% higher than their counterparts. All other symptoms investigated failed to show statistically significant association with the outcome.
Effect modification analyses were conducted to investigate the possible interaction between age and presence of dyspeptic symptoms, age and father education, age and family income, and age and ethnicity over H. pylori prevalence. None of those joint effects showed to be statistically significant in crude or adjusted analyses.
Discussion
The principal limitation of this study is the high refusal rate of the eligible individuals to realization of the 13C-UBT. Despite the simplicity and the non-invasiveness character of the 13C-UBT, which could make it suitable for use in field-based population studies, the refusal rate was very high (36.2%) what, as will be considered in detail below, can compromise the validity of some of the observed results. The need of 6-hour fastening is probably one of the main constraints for its use in field-based surveys. Nevertheless, similar or even higher refusal rates were reported by other authors using different methods in population-based studies collecting primary data: 32% in Australia[19], 41.6% in Northern Ireland[20] and 74% in England[21]. Comprehensively, normal individuals are less compelled to adhere to a test than patients attending consultation for abdominal symptoms, the target population more frequently found in the published literature on H. pylori prevalence. Studies conducted among symptomatic individuals, however, suffer from a different kind of selection bias since findings obtained from samples of gastroenterological patients are not necessarily representative of the whole population and can not be extrapolated with confidence to non-patient individuals. It is not surprising that prevalence rates available in the literature come mainly from the baseline phase of large community interventions focusing on modification of cardio-vascular risk factors and conducted in developed countries [19-21].
Despite the refusals and considering that the prevalence rate of H. pylori infection was over 50% among individuals non-exposed to most of the exposures investigated, the study had a power of 80% to detect relative risks ≥ 2.0, at the significance level of 5%. Lack of power may have impaired the detection of association between the variables abdominal fullness, early satiety, vomiting, and abdominal bloating, because their prevalence was under 30% in the study population.
Another limitation of this study is the impossibility of declaring causality between associated factors and the outcome. The temporality between exposure and outcome can not be ascertained with precision in cross-sectional studies.
To the authors' knowledge, this is the first population-based study of H. pylori infection targeting urban adult population in Brazil. Overall 64.3% of the population ≥ 20 years old was infected with H. pylori. This figure is higher than the prevalence of infection detected in population-based studies conducted in developed countries like Australia (30.6%)[19], England (27.6%)[21] and United States (32.7%)[22], and similar to prevalence rates detected in developing settings in South America, Africa and parts of Asia[1]. Considering that the occurrence of infection was significantly associated to dyspepsia even after allowing for confounding and that refusals were more prevalent among non-symptomatic individuals, it is possible that the self-selective inclusion of symptomatic individuals may have in part overestimated the true prevalence of the infection. Projecting the infection rate observed among symptomatic and non-symptomatic complier subjects to symptomatic and non-symptomatic non-complier participants, however, the prevalence rate would be 64.0% (95% CI 60.0%–68.0%).
In the present study past socio-economic variables in childhood presented the strongest association with the occurrence of the infection in adult life (an adjusted increase of approximately 50% in the probability of infection). The age-dependent increase of H. pylori prevalence which is observed in most of the studies conducted worldwide lost the significance in the present study when adjusted by socio-economic conditions in childhood. This finding is in agreement with the "birth cohort phenomenon" largely described in the literature. Despite the cross-sectional design and the adverse effect of recall bias, at this and at the study conducted by Mendall et al[23] in London, it was possible to explore the association of past socio-economic indicators in childhood with the infection rate in adulthood, a feature generally disregarded in prevalence studies conducted among adult subjects.
The high prevalence rate observed in Pelotas probably reflects infection acquired in childhood and carried on throughout life. In fact, in a population-based study conducted among children from an urban community in north-east Brazil[24], 75.4% were H. pylori positive by the age of 12–14 years. Another study conducted among low socio-economic children attending an outpatient clinic in Belo Horizonte, Brazil, showed that infection occurs early and increases with age[25].
Without a known significant animal or environmental reservoir for human strains of H. pylori, person-to-person contact appears to be the most likely mode of transmission. As a consequence and as others have shown, the number of siblings, particularly the number of older siblings, domestic crowding and living in orphanages are important determinants of the prevalence of H pylori infection[26,27]. The findings of this study confirmed that larger families and higher exposure as in day-care centres was associated with increased prevalence of the infection.
The study of Replogle et al[28] showed a sex difference, higher in men, in the prevalence of infection. The higher prevalence of H. pylori-associated diseases like peptic ulcer and gastric cancer in males supports the hypotheses of a real association. In the present study no statistical association was observed between sex and dyspeptic symptoms in crude or in adjusted analysis. However, since the refusal rate was higher among men, it is not possible to exclude with certainty that men with increased risk of infection had not been lost preferably to others. This bias may have masked a real association between sex and H. pylori infection, if it actually exists.
Among the demographic variables explored, only ethnicity remained independently associated with the infection. After allowing for present and past socio-economic conditions, non-white individuals presented a probability of infection 32% higher than the observed among the whites. In 1992, Malaty et al[29] in a study conducted in the United States with Hispanics matched with blacks and whites for age and socio-economic status, found that the risk of infection was almost identical in Hispanics and blacks and significantly higher than in whites. As suggested by the authors it was probably a reflection of a generation cohort phenomenon related to the generational distance from very low socio-economic status, i.e., the prevalence of H. pylori in Hispanics and blacks is currently lower than that of their parents but higher than that of the white population, which has experienced higher socio-economic status for several generations. Lack of information in the present study regarding generational socio-economic conditions impaired to test this hypothesis.
In the study of McQuillan et al [22] also conducted in the United States showed that race remained statistically associated with the infection after adjustment for socio-economic factors only in low risk groups, what suggests that ethnicity can be a surrogate for other non-explored factors.
Regarding behavioural factors, the majority of recent studies have not found tobacco use or alcohol consumption to be risk factors for H. pylori infection[30]. A study specifically planned to measure whether smoking or consumption of alcohol or coffee was associated with active H. pylori in southwest England concluded that smoking or coffee consumption were not related to active H. pylori infection and that total alcohol consumption was associated with a small, but not statistically significant, decrease in the odds of infection[31]. In the present study no association was found between these exposures and H. pylori infection. These findings however must be seen with caution because changes in lifestyle due to the development of upper gastrointestinal symptoms may mislead the results of cross-sectional studies. In fact, in the present study, changes in alcohol, coffee and mate intakes were reported, respectively, by 35%, 61% and 35% of the participants with dyspeptic symptoms, most of them having reduced the amount or quitted those intakes as an attempt to deal with the symptoms. At the context of the cross-sectional studies, these variables are generally analyzed taking the current status instead of the status before the onset of upper gastrointestinal symptoms as the exposure. A misclassification error moving the association toward the unit is a possible consequence of such approach.
Regarding gastrointestinal complaints only dyspeptic symptoms were significantly associated with the infection. Meta-analyses of trials which have been done in patients with functional (that is, investigated) dyspepsia have shown no benefit from eradication of H. pylori[32]. Since individuals classified as presenting dyspeptic complaints in the current study were not investigated for identifying the cause of their symptoms, they may have functional dyspepsia or diseases such as peptic ulcer or gastro-oesophageal reflux disease. At this particular, a randomised placebo controlled trial conducted in 36 family practices in Canada to determine whether a "test for Helicobacter pylori and treat" strategy improved symptoms in patients with uninvestigated dyspepsia showed a significant symptomatic benefit at 12 months of follow-up[33].
Conclusion
In this study, the prevalence rate of H. pylori infection was high, following the pattern described by others in developing countries. The socio-economic factors in childhood, the ethnicity and the presence of dyspeptic symptoms were independently associated with the occurrence of the infection.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
ISS conceived the study, participated in its design, analysis and coordination, and drafted the manuscript. JB and ASS participated in the design of the study and carried out the isotope analyses. NCJV participated in the design of the study and performed the statistical analysis. CSH, MCB and RDL participated in the design and coordination of the study and helped to draft the manuscript. All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors are thankful to the International Atomic Energy Agency, Vienna, Austria, for providing the 13C-urea used in the Urea Breath Test and for funding isotope measurements of the collected air samples (Project: ARCAL RLA 6042).
==== Refs
Pounder RE Ng D The prevalence of helicobacter pylori infection in different countries Alim Pharmacol Ther 1995 9 33 39
Helicobacter and Cancer Collaborative Group Gastric cancer and Helicobacter pylori: a combined analysis of 12 case control studies nested within prospective cohorts Gut 2001 49 347 353 11511555 10.1136/gut.49.3.347
Rosenstock S Jørgensen T Bonnevie O Andersen L Risk factors for peptic ulcer disease: a population based prospective cohort study comprising 2416 Danish adults Gut 2003 52 186 193 12524398 10.1136/gut.52.2.186
Engel LS Chow W Vaughan TL Gammon MD Risch HA Stanford JL Schoenberg JB Mayne ST Dubrow R Rotterdam H West AB Blaser M Blot WJ Gail MH Fraumeni JF Jr Population attributable risks of esophageal and gastric cancers J Natl Cancer Inst 2003 95 1404 1413 13130116
Lacy BE Rosemore J Helicobacter pylori: Ulcers and more: the beginning of an era J Nutr 2001 131 2789S 2793S 11584108
Bravo LE Mera R Reina JC Pradilla A Alzate A Fontham E Correa P Impact of Helicobacter pylori infection on growth of children: a prospective cohort study J Pediatr Nutr 2003 37 614 619 10.1097/00005176-200311000-00021
Eslick GD Yan P Xia HH Murray H Spurrett B Talley NJ Foetal intrauterine growth restrictions with Helicobacter pylori infection Alim Pharmacol Ther 2002 16 1677 1682 10.1046/j.1365-2036.2002.01333.x
Indicadores e dados básicos-Brasil-2004
Rocha GA Queiroz DM Mendes EN Oliveira AM Moura SB Silva RJ Source of Helicobacter pylori infection: studies in abattoir workers and pigs Am J Gastroenterology 1992 87 1525
Souto FJD Fontes CJF Rocha GA Oliveira AMR Mendes EN Queiroz DMM Prevalence of Helicobacter pylori infection in rural area of the state of Mato Grosso, Brazil Mem Inst Oswaldo Cruz 1998 93 171 174 9698887
Tsugane S Fahey MT Hamada GS Kabuto M Miyakawa VY Helicobacter pylori infection and atrophic gastritis in middle-aged Japanese residents of São Paulo and Lima Int J Epidemiol 1999 28 577 582 10405867 10.1093/ije/28.3.577
Ito LS Oba SM Hamajima N Marie SKN Uno M Shinjo SK Kino A Lavilla F Inoue M Tajima K Tominaga S Helicobacter pylori seropositivity among 963 Japanese Brazilians according to sex, age, generation, and lifestyle factors Jpn J Câncer Res 2001 92 1150 1156
Nascimento RB valente SRG Oliveira LCM Seroprevalence of Helicobacter pylori infection in chronic chagasic patients, and in the rural and urban population from Uberlândia, Minas Gerais, Brazil Rev Inst Méd Trop S Paulo 2002 44 251 254
Lyra AC Santana G Santana N Silvany-Neto A Magalhaes E Pereira EM Mascarenhas R Lyra MC Veiga A Ferreira K Zaterka S Lyra LG Seroprevalence and risk factors associated with Helicobacter pylori infection in blood donnors in Salvador, Northeast-Brazil Braz J Infect Dis 2003 7 339 345 14552744 10.1590/S1413-86702003000500009
Almeida Cunha RP Alves FP Rocha AM Rocha GA Camargo LM Nogueira PO Camargo EP Queiroz DM Prevalence and risk factors asociated with Helicobacter pylori infection in native populations from Brazilian Western Amazon Trans R Trop Med Hyg 2003 97 382 386 10.1016/S0035-9203(03)90063-0
Zubillaga M Oliveri P Panarello H Buzurro M Adami J Goldman C Calmanovici G Alak M Degrossi O Carol R Boccio J Stable isotope techniques for the detection of Helicobacter pylori infection in clinical practice. 13C-urea breath test in different experimental conditions Acta Physiol Pharmacol Ther Latinoam 1999 49 101 107 10797847
Barros AJ Hirakata VW Alternatives for logistic regression in cress-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio BMC Med Res Methodol 2003 3 21 14567763 10.1186/1471-2288-3-21
Victora CG Huttly SR Fuchs SC Olinto MTA The role of conceptual frameworks in epidemiological analysis: a hierarchical approach Int J Epidemiology 1997 26 224 227 10.1093/ije/26.1.224
Peach HG Pearce DC Farish SJ Helicobacter pylori infection in an Australian regional city: prevalence and risk factors M J A 1997 167 310 313
Murray LJ McCrum EE Evans AE Bamford KB Epidemiology of Helicobacter pylori infection among 4742 randomly selected subjects from northern Ireland Int J Epidemiol 1997 26 880 887 9279623 10.1093/ije/26.4.880
Moayyed P Axon ATR Feltbower R Duffett S Crocombe W Braunholtz D Richards ID Dowell AC Forman D Leeds HELP Study Group Relation of adult lifestyle and socioeconomic factors to the prevalence of Helicobacter pylori infection Int J Epidemiol 2002 31 624 631 12055165 10.1093/ije/31.3.624
McQuillan GM Kruszon-Moran D Kottiri BJ Curtin LR Lucas JW Kington RS Racial and ethnic differences in the seroprevalence of 6 infectious diseases in the United States: data from NHANES III, 1988–1994 Am J Public Health 2004 94 1952 1958 15514236
Mendall MA Goggin PM Molineaux N Levy J Toosy T Strachan D Northfield TC Childhood living conditions and Helicobacter pylori seropositivity in adult life Lancet 1992 339 896 897 1348299 10.1016/0140-6736(92)90931-R
Rodrigues MN Queiroz DM Bezerra Filho JG Pontes LK Rodrigues RT Braga LL Prevalence of Helicobacter pylori infection in children from na urban community in north-east Brazil and risk factors for infection Eur J Gastroenterol Hepatol 2004 16 201 205 15075995 10.1097/00042737-200402000-00013
Oliveira AM Queiroz DM Rocha GA Mendes EN Seroprevalence of Helicobacter pylori infection in children of low socioeconomic level in Belo Horizonte, Brazil Am J Gastroenterol 1994 89 2201 2204 7977242
McCallion WA Murray LJ Bailie AG Dalzell AM O'Reilly DP Bamford KB Helicobacter pylori infection in children: relation with current household living conditions Gut 1997 39 18 21 8881801
Malaty HM Paykov V Bycova O Ross A Graham DP Anneger JF Graham DY Helicobacter pylori and socioeconomic factors in Russia Helicobacter 1996 1 82 87 9398883
Replogle ML Glaser SL Hiatt RA Biological sex as a risk factor for Helicobacter pylori infection in health young adults Am J Epidemiol 1995 142 856 863 7572962
Malaty HM Evance DG Evance DJ JrGraham DY Helicobacter pylori in Hispanics: comparison with blacks and whites of similar age and socioeconomic class Gastroenterology 1992 103 813 816 1499931
Brown LM Helicobacter pylori: epidemiology and routes of transmission Epidemiol Rev 2000 22 283 297 11218379
Murray LJ Lane AJ Harvey IM Donovan JL Nair P Harvey RF Inverse relationship between alcohol consumption and active Helicobacter pylori infection: the Bristol Helicobacter project Am J Gastroenterol 2002 97 2750 2755 12425543 10.1111/j.1572-0241.2002.07064.x
Laine L Schoenfeld P Fennerty MB Therapy for Helicobacter pylori in patients with nonulcer dyspepsia. A meta-analysis of randomized, controlled trials Ann Intern Med 2001 134 361 369 11242496
Chiba N Van Zanten SJ Sinclair P Ferguson RA Escobedo S Grace E Treating Helicobacter pylori infection in primary care patients with uninvestigated dyspepsia: the Canadian adult dyspepsia empiric treatment-Helicobacter pylori positive (CADET-Hp) randomised controlled trial BMJ 2002 324 1012 1016 11976244 10.1136/bmj.324.7344.1012
|
16283940
|
PMC1308822
|
CC BY
|
2021-01-04 16:28:59
|
no
|
BMC Public Health. 2005 Nov 10; 5:118
|
utf-8
|
BMC Public Health
| 2,005 |
10.1186/1471-2458-5-118
|
oa_comm
|
==== Front
BMC Public HealthBMC Public Health1471-2458BioMed Central London 1471-2458-5-1201628750010.1186/1471-2458-5-120Research ArticleHispanic physicians' tobacco intervention practices: a cross-sectional survey study Soto Mas Francisco G [email protected] Richard L [email protected] Holly E [email protected] Hsu Chiehwen [email protected] Ximena [email protected] William M [email protected] Department of Social and Behavioral Sciences, School of Public Health, University of North Texas Health Science Center, Fort Worth, Texas, USA2 Department of Health Promotion, University of Nevada Las Vegas, Las Vegas, Nevada, USA3 Department of Kinesiology, Health Promotion & Recreation, University of North Texas, Denton, Texas, USA4 Department of Public and Community Health, University of Maryland College Park, College Park, Maryland, USA5 Department of Physical Performance and Development, Health Education Program, University of New Mexico, Albuquerque, New Mexico, USA2005 14 11 2005 5 120 120 11 6 2005 14 11 2005 Copyright © 2005 Soto Mas et al; licensee BioMed Central Ltd.2005Soto Mas et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
U.S. Hispanic physicians constitute a considerable professional collective, and they may be most suited to attend to the health education needs of the growing U.S. Hispanic population. These educational needs include tobacco use prevention and smoking cessation. However, there is a lack of information on Hispanic physicians' tobacco intervention practices, their level of awareness and use of cessation protocols, and the type of programs that would best address their tobacco training needs. The purpose of this study was to assess the tobacco intervention practices and training needs of Hispanic physicians.
Methods
Data was collected through a validated survey instrument among a cross-sectional sample of self-reported Hispanic physicians. Data analyses included frequencies, descriptive statistics, and factorial analyses of variance.
Results
The response rate was 55.5%. The majority of respondents (73.3%) were middle-age males. Less than half of respondents routinely performed the most basic intervention: asking patients about smoking status (44.4%) and advising smoking patients to quit (42.2%). Twenty-five percent assisted smoking patients by talking to them about the health risks of smoking, providing education materials or referring them to cessation programs. Only 4.4% routinely arranged follow-up visits or phone calls for smoking patients. The majority of respondents (64.4%) indicated that they prescribe cessation treatments to less than 20% of smoking patients. A few (4.4%) routinely used behavioral change techniques or programs. A minority (15.6%) indicated that they routinely ask their patients about exposure to tobacco smoke, and 6.7% assisted patients exposed to secondhand smoke in understanding the health risks associated with environmental tobacco smoke (ETS). The most frequently encountered barriers preventing respondents from intervening with patients who smoke included: time, lack of training, lack of receptivity by patients, and lack of reimbursement by third party payers. There was no significant main effect of type of physician, nor was there an interaction effect (gender by type of physician), on tobacco-related practices.
Conclusion
The results indicate that Hispanic physicians, similarly to U.S. physicians in general, do not meet the level of intervention recommended by health care agencies. The results presented will assist in the development of tobacco training initiatives for Hispanic physicians.
==== Body
Background
Few studies have reported on the particular perspectives and actual practices of U.S. Hispanic physicians with respect to tobacco counseling and smoking cessation. There is a scarcity of information related to how Hispanic physicians perceive their responsibility to educate patients who smoke. In addition, the literature does not clearly discuss their awareness of available resources for assisting smoking patients, nor how tobacco-use assessment, counseling, and follow-up are incorporated into their actual practice. This is of particular importance considering that Hispanic physicians may confront situations that are particular not only to their patient populations, but also to their own cultural and educational backgrounds. In addition, despite the fact that Hispanic physicians are underrepresented nationally, this group constitutes a considerable professional collective. The American Medical Association's membership includes more than 28,400 Hispanic physicians, [1] and the National Hispanic Medical Association represents more than 36,000 licensed Hispanic physicians. [2] In certain states, Hispanic physicians represent a significant percentage of the total physician population: approximately 11% in Florida, 8% in New Mexico and 7% in Texas [3]. Finally, given that smoking continues to be a priority health behavior problem among Hispanics, and that physicians can potentially play an important role in delivering education messages, [4], assessing the tobacco-related training needs of Hispanic physicians should be considered a priority.
The most recent national data show that current smoking prevalence among Hispanic adults is at 16.7%, [5] and that more than 20% of Hispanic high school students are smokers [6]. This constitutes more than 6 million Hispanic youth and adult smokers in need of assistance. Although the literature indicates that tobacco dependence and desire to quit is prevalent across all racial and ethnic groups, and that smoking cessation interventions have shown to be effective in both the general population as well as minority populations, interventions and treatments must be tailored to the cultural characteristics of the participant population [7,8]. Furthermore, information and education must be communicated in a language that is understood by the smoker [7]. There is a recognized need for programs that are specifically developed to address the health needs of Hispanics, including programs for tobacco use prevention and smoking cessation, [8] and Hispanic physicians could play a key role in attending to these tobacco education and smoking cessation needs. According to the literature, race, culture, and language are important factors among Hispanics when selecting their physicians [9-12]. Therefore, if Hispanics prefer physicians who share their same ethnicity, the Hispanic physician may be most suited to attend to the health education needs of this growing population group.
It is essential to better understand how Hispanic physicians perceive their responsibility about educating patients who smoke, and to shed light on their actual practices related to tobacco use assessment, counseling and follow-up. The purpose of this study was to conduct an assessment of the tobacco intervention practices and training needs of Hispanic physicians through a validated survey instrument.
Methods
Data was collected through a validated survey developed by the investigators using qualitative and quantitative approaches. The qualitative study involved the analysis of primary data collected through semi-structured taped interviews with nine practicing Hispanic physicians representing various geographic areas in the state of New Mexico. The study protocol was approved by the University of New Mexico Institutional Review Board. Participants were asked to discuss issues related to tobacco use among their patients as well as their personal perspectives and practices regarding tobacco/smoking intervention. Tapes were transcribed and data compiled using a key word coding system that focused on answering the questions of interest for this study.
The results of the qualitative analysis, along with a review of the related literature, informed the development of the initial draft survey, which was then pilot-tested. To assess face and content validity, the survey was presented to a panel of three health education experts who ensured that all domains of interest were appropriately captured, and to a language specialist, who provided feedback on wording and grammar. A second draft was then developed and piloted with thirteen Hispanic physicians with the purpose of improving the face validity of the instrument by involving the participating population. This data also provided information on the administration process. Most participants completed the survey in 10–13 minutes.
Based on the feedback received, a final instrument was developed and psychometric tests conducted. These included test-retest reliability (N = 8) and internal consistency (N = 45). The results of the test-retest Pearson correlation coefficient (see Table 1) indicated acceptable reliability across the eight items used to assess the domain of interest (see Table 3), as typically seen in validation studies with adults [13]. The Cronbach alpha value was 0.81, which is an acceptable level of reliability. Reliability was estimated by computing the average scores across all the items in the domain to produce a scale score. Thus each item is weighted equally in contributing to the total score.
Table 1 Test-retest coefficients for each item
Item Coefficient
1 .98
2 .90
3 .69
4 .81
5 .66
6 .62
7 .82
8 .92
Table 2 Characteristics of respondents
Item n (%)
Sex
Female 12 26.7
Male 33 73.3
Age group
20–35 1 2.2
36–45 22 48.9
46–50 9 20.0
50+ 13 28.9
Place of birth
USA 41 91.1
Mexico 2 4.4
Puerto Rico 2 4.4
Professional category
Primary care 21 46.7
Specialist 24 53.3
Years of practice in the U.S.
1–3 3 6.7
3–6 1 2.2
6–10 8 17.8
10+ 31 68.9
Type of practice
Private office 23 51.1
HMO 2 4.4
Hospital 7 15.6
Non-hospital based clinic 4 8.9
Country of medical training
USA 42 93.3
Mexico 1 2.2
Puerto Rico 1 2.2
Language spoken at home
English 41 91.1
Spanish 1 2.2
Both 3 6.7
Language most spoken in practice
English 35 77.8
Spanish 2 4.4
Both 8 17.8
Table 3 Physicians' tobacco-related practices in a typical office visit
Intervention Frequency and percent of physicians who perform the intervention with:
>80% of patients 61–80% of patients 41–60% of patients 20–40% of patients <20% of patients
Ask about smoking status (20) 44.4% (10) 22.2% (5) 11.1% (4) 8.9% (6) 13.3%
Advise smoking patients to quit (19) 42.2% (8) 17.8% (5) 11.1% (5) 11.1% (7) 15.6%
Assist smoking patients (11) 24.4% (6) 13.3% (7) 15.6% (4) 8.9% (16) 35.6%
Arrange follow-up for smoking patients (2) 4.4% (4) 8.9% (3) 6.7% (4) 8.9% (30) 66.7%
Prescribe NRT or other treatment (0) 0.0% (0) 0.0% (6) 13.3% (10) 22.2% (29) 64.4%
Use/refer behavioral change programs (2) 4.4% (0) 0.0% (3) 6.7% (7) 15.6% (33) 73.3%
Ask patients about exposure to ETS (7) 15.6% (2) 4.4% (3) 6.7% (11) 24.4% (22) 48.9%
Assist patients exposed to ETS (3) 6.7% (2) 4.4% (5) 11.1% (6) 13.3% (28) 62.2%
The outcome of the described development process was a 50-item instrument that is population-specific and appropriate for assessing the professional characteristics and tobacco practices and educational needs of Hispanic physicians. In addition to standard demographic questions, the instrument includes items related to country of education and professional training, language spoken at home and in professional practice, and ethnicity and language of the patient population. Items related to physicians' smoking status include past and present use of both cigarettes and cigars (the instrument is available upon request).
Eight items assess physicians' tobacco intervention practices, as follows: 1) ask patients about smoking status, 2) advise smoking patients to quit, 3) assist smoking patients by talking to them about the health risks of smoking, providing materials, or referring, 4) arrange follow-up for smoking patients-follow-up visits or phone calls, 5) prescribe Nicotine Replacement Therapy (NRT) or other cessation treatments to smoking patients, 6) use behavioral change techniques with smoking patients, or refer them to behavioral change programs, 7) ask patients about exposure to secondhand smoke, and 8) assist patients exposed to secondhand smoke in understanding the health risks associated with ETS by talking to them or providing materials. Physicians were asked to indicate the percentage of patients with whom they perform each activity in a typical office visit according to the following scale: <20%, 20–40%, 41–60%, 61–80%, and >80% (see Table 3). Greater than 80% was established as the standard for defining "routine practice." This figure was based on the Healthy People 2000 Objectives for the Nation, which established 75% as the benchmark for "routine" tobacco intervention practices. Respondents were also asked to indicate the specific factors which personally prevent them from intervening with patients who smoke, and whether they have access to resources to assist smoking patients. The last section explores tobacco counseling training preferences.
This study used a cross-sectional design that included physicians who were members of the New Mexico Hispanic Medical Society (NMHMS). The NMHMS approved the project and encouraged participation. The unit of analysis and observation for this study consisted of the individual physician. The main selection criterion for inclusion was that participants be practicing physicians. A potential sample population of 81 physicians qualified for the study. A packet including a cover letter, IRB approved informed consent, the survey, and a self-stamped return envelope was mailed to all potential participants. A follow-up letter was sent to non-respondents a month later, and a few weeks later a second packet was mailed to those who had not yet responded. This second packet included a letter offering an incentive to participate (a $15 gift certificate).
Results
Forty-five surveys, or 55.5% of the initial sample size, were entered into the dataset. All data screening, computation, and analyses were conducted using SPSS 10.0 for Microsoft Windows. Characteristics of respondent physicians are included in Table 2. The majority of respondents were male, in the 36–45 years-of-age group, and born in the U.S. The number of participants who were specialists was slightly higher than that of those who were primary care physicians (Primary care physicians generally include family practice and internal medicine. Pediatricians are sometimes considered primary care physicians for children, adolescents and teenagers while obstetricians/gynecologists are sometimes considered primary care physicians for women. For this study, primary care physicians were considered those who responded "Yes" to the question "Primary Care Physician?"). The majority completed medical training in the U.S., had been practicing for more than 10 years, and indicated that they speak English at home and in their professional practice. Although 26.7% (n = 12) of participant physicians had smoked more than 100 cigarettes in their lifetime, none were current cigarette smokers at the time of completing the survey. Only 4.4% (n = 2) had smoked more than 100 cigars in their lifetime; however, among these, 6.7% (n = 3) reported being current cigar smokers.
Information on the patient population of participant physicians was collected. Overall, respondents estimated that they see about the same percentage of Hispanic/Latino and Anglo/Caucasian patients, 6.7% (n = 3) indicated that their patients speak mostly Spanish, and 22.2% (n = 10) that their patients speak both Spanish and English. Forty percent (n = 18) of participant physicians estimated that about one-fourth of their patient population smoke, 35.6% (n = 16) estimated that about one-fourth suffer from tobacco-related illness, and 28.9% (n = 13) estimated that about half or more suffer from tobacco-related illness.
Figure 1 shows the variability in tobacco-related practices among respondents. Less than 45% of participant physicians routinely perform the most basic intervention: asking patients about smoking status and advising smoking patients to quit. Even fewer, 24.4%, routinely assist smoking patients by talking to them about the health risks of smoking, providing education materials or referring them to cessation programs. Only 4.4% routinely arrange follow-up visits or phone calls for smoking patients.
Figure 1 Percentage of physicians performing the 4 A's of smoking cessation Counseling with >80% of patients.
In regard to treating smokers, none of the physicians reported prescribing nicotine replacement therapy or any other type of treatment to more than 60% of smoking patients. To the contrary, the majority (64.4%) indicated that they prescribe cessation treatments to less than 20% of smoking patients. An even smaller percentage, only 4.4%, reported routinely using behavioral change techniques or referring smokers to behavioral change programs. As far as secondhand smoke is concerned, only 15.6% indicated that they routinely ask their patients about exposure to tobacco smoke and 6.7% routinely assist patients exposed to secondhand smoke in understanding the health risks associated with ETS (see Table 3).
Participants were asked to report on the two most frequently encountered barriers preventing them from intervening with patients who smoke by selecting from a list of potential barriers identified during the qualitative study. These barriers were related to available time, reimbursement, training, receptivity of the patient, patient's choice, and language. Of those who reported not always counseling their patients, the most frequently selected barriers included: patients are not receptive (50%, n = 15); counseling takes too much time (33.3%, n = 10); I do not have the proper training to do an intervention (30, n = 9); and counseling time is not reimbursable by third party payers (20%, n = 6). However, 23.3% (n = 7) selected "other" reasons not specified.
Regarding availability and access to tobacco use education and control resources, almost half of respondents (46.6%, n = 21) were not sure about community resources they could use for assisting patients who smoke, and 26.6% (n = 12) indicated that they did not have access to cessation programs or technical support. The majority (73.3%, n = 33) responded either that they were not sure whether they had the materials they needed in their office to adequately educate smokers, or indicated they simply did not have the necessary materials. When asked about resources they would use should they become available, the majority indicated that they would use bilingual Spanish materials, including cessation programs (60.0%, n = 27), education materials (62.2%, n = 28), and quitting self-help guides (64.4%, n = 29).
With respect to tobacco-related training, almost all participants (97.7%, n = 44) indicated a preference for educating themselves about tobacco intervention skills through articles in scientific/professional journals. Written materials, such as manuals, brochures, etc., were the second preferred choice selected by 60% (n = 27) of respondents, while 35.5% (n = 16) selected continuing education courses.
Factorial analyses of variance were conducted to determine whether there were significant mean score differences on the variable of interest by gender, by type of physician (primary care or not), or by the interaction of gender and type of physician. Results indicated a main effect of gender on tobacco intervention practice, with male physicians scoring significantly more positively than female physicians. Male physicians (M = 20.20, SD = 6.50) scored nearly 30% higher than female physicians (M = 14.75, SD = 6.14) (see Table 4). There was no significant main effect related to type of physician, nor was there an interaction effect (gender by type of physician) on tobacco-related practices.
Table 4 Factorial ANOVA results indicating main and interaction effects of gender and type of physician on tobacco-related practices
Source of Variance Sum of Squares df Mean Square F p
Gender 287.89 1 287.89 6.87 .01*
Type of Physician 37.38 1 37.38 .89 .35
Gender by Type of Physician 32.24 1 32.24 .77 .39
Error 1717.12 41 41.88
*R Squared = .152
Discussion
The study provides relevant demographic information related to respondents' smoking status, ethnicity, and language use. Physicians who participated in the study reported both low prevalence of cigarette smoking and low levels of smoking intervention. It should be noted some studies have shown no relation between a physician's own smoking status and tobacco-related practices [14,15]. Regarding the smoking status of the patient population, it is interesting that a high percentage of respondents estimated that about 25% of the patients they see are smokers, a figure which is consistent with the national average. It is even more significant that nearly 30% of respondents estimated that half of their patients suffer from tobacco-related illnesses. This points to the enormous impact of tobacco on health care costs, and suggests that costs could be significantly reduced by eliminating smoking among the study population.
It is important to mention the demographic information related to ethnicity that emerged from the data. Respondents estimated that 47% of their patients are Hispanic, which is higher than the estimated Hispanic population of New Mexico at the time of the study (42%) [16]. Although this figure was estimated by respondents, it would be worth exploring further because it may support the hypothesis that Hispanics prefer physicians who are of their own race/ethnicity. If this is true, the Hispanic physician may be well posited for educating the Hispanic population and delivering tobacco education and smoking cessation interventions to this group.
Results related to tobacco intervention were diverse. Although the qualitative study indicated that participants understand the health burden caused by tobacco use, and feel responsible for intervening with patients who smoke, their actual practices do not reach a level that would significantly contribute to tobacco control. It is significant that less than half of the respondents in this study routinely ask about smoking status, particularly because tobacco use is generally recorded in the medical chart. This was confirmed by the interviews during the qualitative study. Respondents indicated that tobacco use is assessed during intake, and that they know whether a patient uses tobacco by looking at the chart. This may be relevant when taking a systems approach to smoking cessation: simply institutionalizing the recording of smoking status in the medical chart may have a very limited impact on cessation intervention by physicians. Similar conclusions have been reported by Zapka et al [17].
In this study, less than half of respondents reported routinely asking patients about smoking status and advising smoking patients, and less than one fourth routinely talk with patients about the health risks of smoking, provide education materials, or refer them to cessation programs. These results compare negatively with data from national studies, and suggest that the practices of Hispanic physicians may negatively compare with those of their peers. Goldstein et al [15]. assessed community-based primary care physicians and found that 67% "ask" about smoking status during more than 80% of all patient visits; 74% "advise" smoking patients to quit; 35% "assist" smoking patients by talking to them, providing materials, prescribing gum or referring; and 8% "arrange" follow-up visits or phone calls. Similarly, the 2004 State of Health Care Quality report found that more than 68% of current smokers were advised to quit by their practitioners [18].
Whether the lower levels of intervention revealed in this study confirm a national pattern among Hispanic physicians is not clear, given that few national studies have specifically identified Hispanic physicians' level of smoking intervention. However, some studies have found differences in smoking intervention between Hispanic physicians and physicians from other ethnic/racial groups. A study with prenatal, pediatric and WIC providers in greater Boston, Mass., found lower smoking intervention performance among Hispanic providers in comparison with Non-Hispanic Black and White providers, although differences were not significant when controlling for other variables [17].
Regarding ETS, few participants in this study routinely address secondhand smoke in their daily practice. The literature does not provide much information on ETS assessment and counseling by Hispanic physicians, but it is important to note that they may be in a unique position to address the problem. Strong family ties constitute a traditional Hispanic value, and smoking patients may be more willing to follow the advice of a physician not to smoke at home or around family members when it is presented within the context of familismo or family.
The principal barriers to smoking cessation intervention identified in this study include: perceived lack of receptivity by patients, time, lack of training, and lack of economic reimbursement for counseling. These barriers are similar to those found in other studies [17,19,20]. These results indicate that both systematic and individual approaches need to be incorporated into training programs.
Interventions should also consider whether physicians have access to resources for assisting smoking patients. In this study, almost half of the participants indicated that they were not sure about the availability of community resources for smoking cessation, one fourth felt they did not have access to programs and support, and three-fourths were simply not sure or did not have in-office access to all of the materials they needed for educating smokers. Most participants indicated the need for bilingual materials and cessation programs. The qualitative study supported these results. None of the physicians who were interviewed knew about the state tobacco control program in NM.
Finally, almost all respondents indicated a preference for educating themselves about interventions and skills through articles published in scientific/professional journals. This is a significant finding, as most health professionals subscribe to scientific journals. While research papers do not generally discuss the practical applicability of the results, journals interested in contributing to educating professionals on tobacco intervention should require authors to discuss the link between research and practice. The second preferred choice was "other written materials" (brochures, pamphlets, etc.). Only 37.7% (n = 17) selected continuing education courses.
Limitations
A number of limitations of this study should be mentioned. First, the sample size was small, although it represents approximately 15% of the Hispanic physician population in NM [3]. Additionally, all participants were members of a professional organization, which may have influenced their responses. In addition, survey data were self-reported, and, although the survey instrument demonstrated good validity and reliability, providers' self-reported practices may be less valid than data obtained from other sources [23]. Additionally, self-reported practices may be more positive among respondents than among non-respondents. Finally, due to the small sample size of the study, results must be interpreted cautiously when generalizing to the general Hispanic physician population. Considering these limitations, the implementation of a follow-up national study with a larger sample and a more diverse participant population is recommended. It would be useful to compare the perceptions, opinions, and practices of Hispanic physicians of diverse nationalities and educational backgrounds. A larger and more comprehensive study would provide valuable information that could potentially be used to inform the development of interventions to educate Hispanic physicians about tobacco assessment, counseling, and follow-up.
Conclusion
To the knowledge of these investigators, this is the most comprehensive study that has explored the tobacco intervention practices of U.S. Hispanic physicians. In addition to the results of the qualitative and quantitative studies, another positive outcome of this investigation was the development of a survey instrument that showed good validity and reliability. This instrument should prove to be of interest to federal health agencies and managed care organizations, for investigating Hispanic physicians' tobacco intervention practices. According to the literature, standard procedures for collecting data on physicians' tobacco education and smoking cessation practices present several methodological challenges. These include the use of survey instruments that have not undergone the necessary processes to demonstrate their validity and reliability. The use of such instruments may compromise the internal validity and results of an investigation [21,22]. Furthermore, the instrument developed for this study includes questions not generally found in other surveys which make the instrument more appropriate for assessing the characteristics of Hispanic physicians, including items related to level of acculturation, language proficiency and training needs.
In summary, although the literature has consistently reported on physicians' low level of compliance with smoking cessation guidelines and recommendations, this study suggests that Hispanic physicians may be in greater need of training and resources than other groups. The results of this study compare negatively with data from national studies on physicians in general. Regarding "best practices" for delivering training initiatives to Hispanic physicians, the results of this study point to self-education through professional/scientific journals. However, this result must be further investigated. It could be that Hispanic physicians prefer to learn about tobacco intervention by reading scientific journals, but it may turn out that this approach is not conducive to increased tobacco intervention. Connecting research and practice is an issue of concern in health education, and journal articles will certainly not provide adequate training if researchers do not explain the practical applicability of their research findings.
Given the growth of the Hispanic population in the U.S., and the demonstrated potential role of physicians in smoking cessation, more resources should be dedicated to identifying the tobacco intervention needs of Hispanic physicians, and to improving their cessation practices.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
FGSM designed the study, conducted data collection and analysis, and wrote the first draft of the manuscript. RLP contributed to study design, data analysis, interpretation of results, and technical and editorial review. HEJ contributed to study design and editorial review. CEH contributed to data analysis and manuscript development, including technical and editorial review. XU-R contributed to interpretation of results and technical and editorial review. WMK contributed to study design, interpretation of results and technical review.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors would like to give special recognition to Dr. William Kane for his many years of dedication to higher education and to health education.
This study was supported by the University of New Mexico Graduate College through a General Grant Award to the first author.
==== Refs
Total physicians by race/ethnicity-2003 2005 Physician characteristics and Distribution Department of Data Quality and Measurment, Division of Data and Operations, American Medical Association, 2004
National Hispanic Medical Association Accessed May 12, 2005
Distribution of nonfederal physicians by race, 2003 Accessed May 9, 2005
Soto Mas F Papenfuss R Medicina y salud pública Médico Interamericano 1997 16 206 209
Adult cigarette smoking in the United States: current estimates. May, 2004 Accessed December 7, 2004
Youth and tobacco use: current estimates. December, 2003 Accessed December 7, 2004
DHHS Treating tobacco use and dependence. Clinical Practice Guideline 2000 U.S. Department of Health and Human Services. Public Health Service. Rockville, MD: Author
DHHS Healthy People 2010: understanding and improving health 2000 U.S. Department of Health and Human Services. Washington, DC: U.S. Government Printing Office
Saha S Taggart SH Komaromy M Bindman AB Do patients choose physicians of their own race? Health Aff 2000 19 76 83 10.1377/hlthaff.19.4.76
Cooper-Patrick L Gallo JJ Gonzales JJ Vu HT Powe NR Nelson C Ford D Race, gender, and partnership in the patient-physician relationship JAMA 1999 282 583 589 10450723 10.1001/jama.282.6.583
Morales L Cunningham W Brown J Hays R Are Latinos less satisfied with communication by health care providers? J Gen Intern Med 1999 14 409 417 10417598 10.1046/j.1525-1497.1999.06198.x
David R Rhee M The impact of language as a barrier to effective health care in an underserved urban Hispanic community Mt Sinai J Med 1998 65 393 397 9844369
Willet W Nutritional epidemiology 1990 New York, NY: Oxford University Press 92 126
Saywell RM Jay SJ Lukas PJ Casebeer LL Mybeck KC Parchman ML Haley AJ Indiana family physicians attitudes and practices concerning smoking cessation Indiana Medicine 1996 March/April 149 156 8867414
Goldstein MG DePue JD Monroe AD Lessne CW Rakowski W Prokhorov A Niaura R Dube CE A population-based survey of physician smoking cessation counseling practices Prev Med 1998 27 720 729 9808804 10.1006/pmed.1998.0350
New Mexico QuickFacts Accessed September 20, 2005
Zapka JG Pbert L Stoddard AM Ockene JK Valentine Goins K Bonollo D Smoking cessation counseling with pregnant and postpartum women: a survey of community health centers Am J Public Health 2000 90 78 84 10630141
The State of Health Care Quality 2004. NCQA\Publications\SOHC 2004 Accessed April 25, 2005
Cabana MD Rand CS Powe NR Wu AW Wilson MH Abboud PA Rubin HR Why don't physicians follow clinical practice guidelines? JAMA 1999 282 1458 1464 10535437 10.1001/jama.282.15.1458
Jaen C Crabtree B Zyzansky S Goodwin M Stange K Making time for smoking cessation counseling J Fam Pract 1998 46 425 428 9598001
Crooks C Kenney E Elder J Levitz M Johnson M Bal D Tobacco Control activities of primary care physicians in California Eval Health Prof 1993 16 311 321
Zapka JG Fletcher KE Ma Y Pbert L Physicians and smoking cessation: development of survey measures Eval Health Prof 1997 20 407 427 10183332
Montano DE Phillips WR Cancer screening by primary care physicians: a comparison of rates obtained from physician self-report, patient survey, and chart audit Am J Public Health 1995 85 795 800 7762712
|
16287500
|
PMC1308823
|
CC BY
|
2021-01-04 16:28:58
|
no
|
BMC Public Health. 2005 Nov 14; 5:120
|
utf-8
|
BMC Public Health
| 2,005 |
10.1186/1471-2458-5-120
|
oa_comm
|
==== Front
BMC PsychiatryBMC Psychiatry1471-244XBioMed Central London 1471-244X-5-431628008810.1186/1471-244X-5-43Research ArticleExperiences in applying skills learned in a mental health first aid training course: a qualitative study of participants' stories Jorm Anthony F [email protected] Betty A [email protected] Stephen K [email protected] ORYGEN Research Centre, University of Melbourne, Locked Bag 10, Parkville, Victoria 3052, Australia2 Centre for Mental Health Research, Australian National University, Canberra, ACT 0200, Australia3 Qualitative and Quantitative Social Research, Gungahlin, ACT 2912, Australia2005 9 11 2005 5 43 43 26 4 2005 9 11 2005 Copyright © 2005 Jorm et al; licensee BioMed Central Ltd.2005Jorm et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Given the high prevalence of mental disorders and the comparatively low rate of professional help-seeking, it is useful for members of the public to have some skills in how to assist people developing mental disorders. A Mental Health First Aid course has been developed to provide these skills. Two randomized controlled trials of this course have shown positive effects on participants' knowledge, attitudes and behavior. However, these trials have provided limited data on participants' subsequent experiences in providing first aid. To remedy this, a study was carried out gathering stories from participants in one of the trials, 19–21 months post-training.
Methods
Former course participants were contacted and sent a questionnaire either by post or via the internet. Responses were received from 94 out of the 131 trainees who were contacted. The questionnaire asked about whether the participant had experienced a post-training situation where someone appeared to have a mental health problem and, if so, asked questions about that experience.
Results
Post-training experiences were reported by 78% of respondents. Five key points emerged from the qualitative data: (1) the majority of respondents had had some direct experience of a situation where mental health issues were salient and the course enabled them to take steps that led to better effects than otherwise might have been the case; (2) positive effects were experienced in terms of increased empathy and confidence, as well as being better able to handle crises; (3) the positive effects were experienced by a wide range of people with varied expectations and needs; (4) there was no evidence of people over-reaching themselves because of over-confidence and (5) those who attended were able to identify quite specific benefits and many thought the course not only very useful, but were keen to see it repeated and extended.
Conclusion
The qualitative data confirm that most members of the public who receive Mental Health First Aid training subsequently provide support to people with mental health problems and that this support generally has positive effects.
==== Body
Background
Community surveys in a range of countries show a high prevalence of mental disorders, but with many cases not receiving appropriate professional help [1]. As well as this unmet need with clinical disorders, there is the problem of how to help the large number of people with sub-clinical symptoms. These people are a source of considerable disability in the population and have a high risk of progressing on to clinical disorders [2]. The problem of unmet need is not uniform within the community. Some sub-groups, such as people living in rural areas, have even lower access to services [3,4].
In order to help overcome the problems of high prevalence, unmet need for treatment, the impact of sub-clinical symptoms, and the uneven geographic distribution of services, there is a need to consider solutions that lie outside formal health services. One approach is to widen the base of people with some knowledge and skills in helping people with mental health problems. Examples of such an approach, which have either been implemented or proposed, include the use of lay people as counsellors in telephone help-lines [5], web sites giving mental health information or therapy directly to the public [5], training of community members, including parents, in how to assist people in crisis situations [6,7], and training of human service providers in rural areas in how to support people suffering from mental disorders [8].
An approach of this sort, which has recently been developed in Australia and has spread to a number of other countries, is Mental Health First Aid (MHFA) training [9,10]. The philosophy behind this approach is that people with mental health problems can potentially be assisted by those in their social network. There is evidence that better social support reduces risk of developing mental disorders and improves outcomes for people experiencing disorders [11,12]. However, potential helpers often lack the confidence and skills to provide basic help. Supporting this view, a recent national survey of the Australian public found that many people had a less than adequate range of first aid responses to people with mental disorders [13]. The MHFA course follows the model that has been successfully applied with first aid for physical disorders. It trains members of the public to give early help to people with developing mental health problems and to give assistance in mental health crisis situations.
Two controlled trials have been carried out to evaluate the MHFA course. The first of these involved 301 employees of government departments who were randomized to either receive the training immediately or to be wait-listed for 5 months before undertaking the training [14]. The trained group improved more than the wait-list control group in terms of confidence in providing help to others, likelihood of advising people to seek professional help, concordance with health professionals about treatments, stigmatizing attitudes, and mental health of the participants themselves.
The second controlled trial was carried out with 753 members of the public in a large rural area of south-east Australia [15]. In this trial, the catchment of an area health service was divided into 16 local government areas. Eight of these areas received the course immediately and the other eight were controls and placed on a waiting list to receive the training later in the year. Relative to controls, people who did the course showed better recognition of disorders from case descriptions, decreased social distance towards people with mental disorders, became more like health professionals in their beliefs about what treatments are likely to be helpful, had greater confidence in providing help to someone, and were more likely to actually provide help.
While these controlled trials showed benefits of the training, the quantitative measures did not seem to fully capture the effects. There have been many instances where participants have told stories, sometimes dramatic ones, of how they used their skills following the course to help someone. It may be more natural for participants to report their experiences in the form of stories, because that is the way they are organized in memory [16], than to answer more abstract questions about them. Furthermore, the quantitative measures did not capture the effects on the recipients of first aid, whether positive or negative, because it was not possible to identify and survey these recipients. By analyzing the stories it is possible to get some indirect information about the effects on the recipients. We therefore sought to gather such stories in a systematic way using participants from the second controlled trial and to analyze the major themes that emerged.
Methods
Sample
The sample was drawn from 131 participants in a cluster randomized trial (ISRCTN53887541) [15]. These were the participants who were in the intervention group (N = 416) and were known to have completed every session of the 9-hour training course in March-April 2003. Those excluded either did not attend the course after random assignment (n = 76), only attended some sessions (n = 33) or had missing attendance data (n = 176). The participants were phoned by one of two research assistants who asked them to complete an additional questionnaire. Out of the 131 approached, 94 actually completed the questionnaire. The questionnaires were completed in November 2004–January 2005, which was 19–21 months after finishing the training course.
Questionnaire
A survey questionnaire was developed to achieve the following: (1) Gather data on respondents with regard to age, gender and education; (2) Ascertain whether the respondent had, or had not, experienced a post-course situation where someone seemed to have a mental health problem; (3) For those who had experienced such a situation, a variety of questions were asked which allowed the respondent to describe that experience; and (4) For those who had not had such an experience, the questionnaire asked them a few questions about expectation and about confidence with such a situation if it were to arise in the future.
For those who had experienced a post-course situation where someone seemed to have a mental health problem, the questions ran as follows: Could you tell us something about the situation(s) and the problems you believed the person(s) was experiencing? Were you able to do anything specific to help the person (s) you believed was suffering the mental health problem(s)? [For those who could not help] What was the reason(s) that you were not able to help that person(s)? Can you give us any examples of something you did? What do you think were the effects on that person/s of what you did? Can you give us any examples of how your relations with that person/s, or your feelings towards them, have changed? Do you think this change had any effect on the person/s, either good or bad? How (if at all) has doing the MHFA course changed how you relate to or feel about the person(s) suffering from that mental health problem?
For those who had not experienced a post-course situation where someone seemed to have a mental health problem, the questions ran as follows: Is this what you would have expected or is it somewhat surprising not to have come across such a situation? If you were now to come across someone who you believed was suffering from a mental health problem, how well prepared would you feel to deal with that? How has doing the MHFA course changed how you relate to or feel about a person(s) suffering from mental health problems? Is there anything else you would like to say about the MHFA course and its value for you?
Survey procedure
Participants were given the option of completing the questionnaire either on-line or on paper. If the former, the participant was asked to provide an email address and was subsequently emailed the URL which gave the questionnaire at a WebSurveyor™ site. By clicking on the URL, the respondent was connected to an on-line facility. This site allows built in switching and skipping, so that only the questions applicable to a given respondent appear on the screen. S/he either selected a button based response for closed-ended questions or typed in free text for open-ended responses. The data generated were downloaded as Excel spreadsheets. Participants who chose to do a paper questionnaire had it mailed out. The paper responses were entered into Excel and the two data sources combined into one. Participants who failed to complete a questionnaire were given a reminder either by email or post, as appropriate.
Analysis
To give independence in the evaluation from the developers of the Mental Health First Aid course, the company Qualitative and Quantitative Social Research was contracted to develop the questionnaire and carry out the analysis. The analysis consisted of three parts: (1) description of the respondents (2) quantitative analysis of frequencies of response and cross-tabulations of respondent characteristics with types of response, and (3) qualitative analysis identifying themes behind responses to each question and selecting stories that illustrate the range of experiences reported. The raw material of the stories was examined in print form. First, the material was read through by one of the authors (SKM) and broad themes were developed. Each major element of an answer was then tagged with a code to identify the themes and the material sorted to ensure that any themes that were identified in the analysis and any quotes that were to be used to illustrate these themes were genuinely representative of the material. Particular emphasis was laid upon ensuring that those themes that emerged frequently were highlighted in the reported analysis. As far as possible, stories quoted are in the exact words of the respondents – some very minimal editing has been undertaken to improve the flow of the story and to tidy up grammar, syntax and spelling.
Results
Characteristics of the respondents
There were 94 respondents to the questionnaire, with 56 responding by mailing a paper copy and 38 completing it on-line. The respondents were predominantly female (75/94) and had an age range from 21–74 years (mean 51 years). With regard to education, 31 had a university degree. Looking at previous experience with mental illness, 29 reported a personal experience, 21 a family experience, 8 reported both, and 36 neither. Compared to the full sample participating in the intervention group of the trial [15], these respondents tended to be better educated (33% vs 21% with a degree), but were similar in age and gender.
Quantitative findings
Table 1 shows the distribution of responses to the basic questions asked. Most respondents had experienced, post-course, a situation where someone seemed to have a mental health problem. Of the 73 who had experienced a mental health related situation, 79% said they had definitely been able to help, 17% said they thought they had been helpful and only 7% were unsure. Of those who had not clearly had any encounter with a mental health issue post course, most felt they could cope well or moderately well with such a situation if it arose.
Table 1 Frequencies of responses to questions
Question Response N with Response
Experienced a post-course situation involving mental health problem Yes 73
No, Not Sure, No Response 21
Total 94
If had an experience: able to help? Yes, definitely 44
Yes, I think I was helpful 10
Unsure 4
No 0
No response 15
Total 73
If not had an experience: surprised? Yes 2
No 9
Don't know 7
No response 3
Total 21
If not had an experience: feel prepared now? Cope well 9
Cope moderately well 12
Unsure or don't know 1
Total 22
The data were also examined to see if there were any apparent patterns of association, either between independent variables, such as age, sex, education or form of response (on paper vs. on line) or between independent and dependent variables (qualities of the experience). The principal method used was to compile contingency tables and then examine various measures of association – as well as visual inspection – as a guide. Formal chi-square tests were not used because cluster effects in the sample violated the assumption of independent observations. Associations are not reported here, because only one association of any apparent size emerged: an association between levels of reported education and how the survey was completed, with on-line completion being associated with higher levels of education. However, the lack of other associations shows that the MHFA course was equally valuable to a wide range of people, irrespective of age, gender and education.
Qualitative findings
Answers to each of the open-ended questions could be analysed as a block of data in their own right. However, it was also possible to connect the responses from consecutive questions to create short stories that illustrate the value of the course.
Stories about first aid
Table 2 presents three stories that are focused mainly around problems that took place within the intimate social network of friends and family, while Table 3 presents three that examine issues that arose in the work setting (broadly defined).
Table 2 Stories of providing first aid that involve the intimate network
Story 1
The situation: [I was] living with a girl with manic depression.
What you did: I recommended some professional help
Effects on that person: She is now on medication and getting on well with work (etc) again.
How relations changed: We no longer live together, she has moved away from where her problems essentially evolved. I now see her in a different light: she was once a fun-loving person, now she rarely leaves home.
Longer term effects on the person: She says now that she is much happier than before.
How the course has changed you: With the MHFA course and my social science degree, I understand the way people act and behave and why they do so.
Anything else: It gave me a basic understanding and some valuable, simple information to help me get hold of the concepts of mental health.
Story 2
The situation: My daughter was suffering from drugs and stress and had a psychotic episode.
What you did: I went to counselling with her and to see the psychiatrist with her. In the end she wanted to be on her own with nothing to do with my husband or me and we "backed off". I did keep in contact via fax: it was one sided but made it easier to keep the dialogue alive.
Effects on that person: My daughter made contact again of her own accord and I believe it was because I never broke contact with her.
How relations changed: I don't think we will ever have the same close relationship as before but we do talk on the phone and see each other at least once a month. I have tried and continue to try not to be too protective, which is what she felt I was.
Longer term effects on the person: I think my daughter has realised that we love her, she is no longer on drugs and I believe that she no longer has the stress she had before.
How the course has changed you: Doing the MHFA made me realise that we were not alone and that there were lots of other people in the same boat. So I did not feel so guilty and it helped my husband and me as well as my daughter knowing we were going to the course.
Anything else: I think it is, no, I know it is an excellent course that has helped so many people and I hope MHFA course will always continue.
Story 3
The situation: [The man concerned was experiencing] severe depression and anxiety due to marriage/family break-up and child custody problems.
What you did: I persuaded him to seek counselling. I tried to listen and advise and I also gave essential financial assistance.
Effects on that person: He did seek counselling and also presented at a mental health unit as an in-patient for a few days but has since discontinued regular counselling.
How relations changed: I now have a little more understanding but I feel there is still an underlying serious problem and I find it difficult to know how to best handle potentially explosive situations.
Longer term effects on the person: There was a temporary effect for the good. He may also be more willing to seek help in the future if needed now he has experienced what the mental health unit can offer.
How the course has changed you: I am somewhat more understanding and make more allowance for irrational behaviour, etc, but it is still not always easy.
Anything else: I am very pleased I did the course and it has made me aware not only of the problems people have due to mental health, but of the help that is available if only the person will seek it.
Table 3 Stories of providing first aid that involve the work setting.
Story 4
The situation: [I'm currently] working with a family who lost a child due to an accident. 'Mum' has had a history of post natal depression and is now experiencing an episode of depression that has lasted in excess of six months. She is taking medication but sometimes cannot afford to get it.
What you did: I offered non-discriminatory, non-judgmental support. I made gentle suggestions about ways she might get child to school, helped her get a Christmas tree up (for the kids) which alleviated her guilt around not feeling like having a Christmas and so on.
Effects on that person: I think she was very grateful that someone seemed to understand and was willing to help her achieve little things rather than expect her to "get her act together." I have become someone who she doesn't have to put on an act for or pretend she isn't home.
How relations changed: Previously when people tried to 'help' she would not answer the phone or pretend she wasn't there. This has not completely changed: recently I visited at an agreed time and no-one answered the door. I could hear music and wondered whether she was in fact there but didn't open the door.
Longer term effects on the person: I feel that we have developed a relationship that is based on gaining small positive outcomes. She is a long, long way from where she would like to be, but we now have a dialogue based on her being able to tell someone how she is feeling and having someone listen.
How the course has changed you: It has made me aware of the signs and symptoms of depression. I have been able to work from where the client is at rather than having specific expectations of their contribution to assistance.
Anything else: I believe that the MHFA course has been particularly valuable to me. I am not a trained counsellor or therapist but in my work with families whose children have disabilities it is likely that some of my families will suffer from depression from time to time.
Story 5
The situation: I work for a Job Network Provider. Many of the long term unemployed suffer from depression and other mental health problems. One example a client was always injuring himself on jobs and on the latest job (the employer said) he wouldn't stop crying. There was also another female client who was suffering from Crohn's Disease who had always been very nervous and worried.
What you did: I referred the woman with Crohn's disease to a holistic GP (as her previous GP did nothing). This new GP referred her to counselling, got her on anti depressants and is looking after her health.
Effects on that person: I saw the difference in this person and she felt the difference of feeling better and learning to cope with life stressors. Last time I saw her she had improved a lot.
How relations changed: I have always been a naturally caring person. I care for that client and most of my clients.
Longer term effects on the person: Good! She not only was getting better but realised she can make a difference to her own life by using the appropriate skills and getting help. She also realised that there are people who care and she was not alone.
How the course has changed you: It confirmed my previous knowledge of Mental Health and gave me a few more ideas of what to do when someone needs help.
Anything else: I think it should run regularly as the population needs to learn how to deal with these sorts of situations and it creates an awareness and perhaps takes away the fear of the unknown.
Story 6
The situation: I was working for a charity organisation for work experience to achieve a Diploma of Community Welfare. I was an interviewer for people of low income or on benefits, when a client entered the room (I was not allowed to participate but just sat there and listen to what the other two interviewers said). This client spoke very fast and told us about the people who he lived with burning his clothes in the lounge room in the middle of the floor. The other two interviewers did not know how to react in this situation: I could not control myself I had to speak up. I just asked him a simple question as whether he was taking medication, he said he wasn't. Then the other two interviewers asked about his doctors, etc.
What you did: By just asking a simple question about medication made the other interviewers realise he had a mental illness.
Effects on that person: He told me that he didn't need his medication any more and he was no longer ill. However, I think that he must have gone to the doctor because I saw him a few weeks later in a different situation and he was acting quite normally.
How relations changed: I really do not think they have changed as I only know him as a client and I have a good understanding towards people I interview. The only change would be the improvement because of medication.
Longer term effects on the person: The change would be for the better because of medication.
How the course has changed you: There has always been a stigma attached to mental illnesses, it has opened my eyes about how I and others feel toward these people. I am more understanding.
Anything else: Most people do not have an understanding of mental illness. If more people could participate in the Mental Health First Aid course it would be a better world. I think the course was a great value to me personally.
In Table 2, Story 1 shows how the course helped someone cope with a challenge, reacting constructively to a partner with a mental disorder. The story is not one with a conventional happy ending – they did not in fact "live happily ever after" – but it does illustrate how the respondent was able to act constructively and make sense for himself about what happened.
Story 2 is somewhat similar in that it centres on dealing with a person with a mental disorder in an intimate, family relationship. Again there is a benefit with regard to increased understanding and self-knowledge. Here, however, we also see that it enabled the respondent to take actions that preserved a relationship that otherwise was in serious jeopardy.
Story 3 is slightly less intimate, but concerns the way a person responded to a friend who was experiencing mental health issues. The main benefits here are focused around the short term – it helped to get the man concerned to get counselling – and in palliating the emotional impact of the challenge. While the respondent does not see that having been at the course solved the problem, clearly he finds the short term and palliative effects valuable.
The three workplace stories (Table 3) are quite different in context, even though they share some themes: the first comes from a person who works directly in a helping profession, the second from someone who works in a context where clients have ancillary problems, while the last example comes from someone who was carrying out work experience.
In Story 4, the respondent clearly identifies benefits for him and for a client with whom he was working who experienced mental health issues. While he does not feel he fully solved anything, he is clearly able to show how deploying the skills he had acquired led to a positive set of developments and small wins.
In Story 5, a person who works with socially disadvantaged clients mentions that he sees many people with mental health issues. Choosing one recent case as an example, he shows how increased understanding and awareness led to taking positive steps that benefited the client.
Story 6 introduces a quite different aspect. Here we see that while the intervention the respondent took was positive for a client, and also increased the respondent's self-awareness and positive feelings, a major benefit is that the knowledge acquired on the course positively role modelled something for co-workers and helped them meet a challenge.
Analysis of themes in open-ended questions
In this section, each open ended-question is presented and the results analysed and illustrated with examples.
Could you tell us something about the situation(s) and the problems you believed the person(s) was experiencing?
As the following illustrations show, some people took the course because of existing problems with friends and family in their life, others because voluntary work brought them into contact with people experiencing mental health problems and there were also quite a few people with professional health care positions who sought to receive additional training. Problems also varied in character from the more serious (schizophrenia) to somewhat less serious (panic attacks) and from the chronic to the episodic.
My son has for many years experienced depression and more recently possibly some kind of psychosis. He admitted himself into a mental health hospital and received help from psychologists on release. He is currently taking anti-depression medication, but I was not sure he would continue to take the drugs.
I was with a friend who had an anxiety attack which included difficulty in breathing, racing heart, chest pains, cold clammy skin and obvious distress.
As I am a trained nurse and work in a Day Room. We have quite a lot of patients with problems. I found the course very informative as my training 40 years ago lacked in psychiatry. My knowledge comes from life and reading and nowadays I can recognise panic attacks, depression and look after these much better.
The large majority of respondents felt they were able to be helpful with the situation faced. Those who could not help were asked "What was the reason(s) that you were not able to help that person(s)?" The three responses to this question referred to the fact that the person was already getting help or did not want help.
Can you give us any examples of something you did?
There was a wide variety of response here. While each response is unique, several broad themes emerged. The foremost theme, especially common for problems with family and friends, was calming the distressed person and listening to them. A sub-theme concerned listening and also actively referring for more specialist assistance.
I helped calm the person, kept them safe with supportive doctor and family members. I also supported without agreeing with their delusions and thoughts.
Listened. Identified that I could understand somewhat of how they saw things. Encouraged to consult a doctor or psychologist.
Two other major themes ran a roughly equal second place. One concerned referral to specialist help, which was more obvious among respondents where the contact was through a professional, workplace situation.
Persuade the person to seek counselling. Tried to listen and advise, also gave essential financial assistance.
The other main theme concerned giving concrete support and practical help and information/advice.
Offered non-discriminatory, non-judgmental support. Made gentle suggestions about ways she might get child to school, helped her get a Christmas tree up (for the kids) which alleviated her guilt around not feeling like having a Christmas.
What do you think were the effects on that person/s of what you did?
A few people responded to this question with general, positive statements such as "helpful" or "excellent" effects. The largest group, however, described immediate impacts, which involved calming, comforting or emotional support.
A calm and relaxed atmosphere brought the client a reassurance that I was not a threat to him, so calmed the client down so he could talk and get what was in his mind out and not be prejudged.
The second largest group overall described long term changes. These included getting the person on medication, giving practical help and establishing better relationships.
Kept him alive during the first severe stages of his illness (about one month), reassurance, gave him the time and space to recover himself, cut to a small degree the despair of "utter aloneness", did a lot of simple things (phone calls, tax, Centrelink, bills, forms etc) that for the duration of the illness were impossible for him but helped keep his "life on track".
It cannot be ascertained for certain, but one sensible interpretation of the contrast between these two groups is that more focus was placed on the immediate effects when the respondent was less heavily involved with the person experiencing the mental health problem and/or the problem was less severe. Where the problem was more severe, and especially when the respondent was more closely involved (e.g. friend or family member) more detailed answers, with a longer time frame, were supplied.
Can you give us any examples of how your relations with that person/s, or your feelings towards them, have changed?
The overwhelming trend in the data here was in the direction of very positive answers, with a relative trickle of answers that were either mixed or negative. Within the positive trend, two sub-trends could be identified, reflecting the double- barrelled character of the question. In the first, the respondent concentrated mainly upon how the relationship improved:
Our relationship has improved. Communication is better and violent mood swings are not as frequent.
The second theme, of course, focused more on positive improvement in the way the respondent felt or thought.
More tolerant – I'm less inclined to get "cross" about my time being "wasted".
A few respondents gave a negative response, although none of these suggested that the course was in any way deficient or that the things they learned failed to work.
Not a lot. Perhaps they are more willing to communicate their feelings.
Clearly, the experience of the course had positive effects for the large majority of people, either in improving their relationship with someone experiencing a mental health problem or in their feelings towards and ideas about that person, or both.
Importantly, here and elsewhere in the analysis, no hint emerged that the course led people into an unrealistic position. For example, no one told a story that suggested they became over-confident and hence made a mess of a situation.
Do you think this change had any effect on the person/s, either good or bad?
The overwhelming response here was that effects had been positive. In a number of cases the respondent simply said "good", or "positive", but many others elaborated.
I believe the change has had a good effect, she is gradually moving forward to make things better in her life with the support of services, GP, etc.
This change has been very positive on both of us. I have seen first hand the benefits of being a good listener and friend and giving helpful advice. My friend states that she is more patient and understanding of others.
Very few overall gave negative responses. Clearly, the dominant response to this question was that the knowledge and skills assisted people to take a constructive and positive response which led to desirable and welcome outcomes.
How (if at all) has doing the MHFA course changed how you relate to or feel about the person(s) suffering from that mental health problem?
Two main themes emerged in the answers here – competence and empathy – both with sub-themes. The largest single group of answers concerned professional competence. A substantial number of people who did the course worked with people in the broad area of health or the helping professions. For these respondents, the course clearly delivered a greatly increased sense of comfort and confidence.
It has opened my eyes to the wide range of mental health problems people may suffer. In my work I deal with unemployed people who may be suffering from mental health problem as a result of unemployment or it may be affecting their ability to hold down employment.
While professional competence was important, a linked but smaller theme was personal competence, that is, the capacity to deal confidently with something that had previously been an on-going challenge.
The course provided me the gateway to education about mental health issues, it has made me more self aware and prepared me better for the experiences that I have had thus far especially as it was emotionally hard for me to accept that a close family member was ill.
The second large theme in response to this question centred upon having developed empathy and understanding. As with competence, this theme sub-divided, in this case almost equally, into two themes: generalised empathy towards people with a mental illness and specific empathy towards someone with whom there had previously been tensions and strain.
I have deeper understanding on what these people actually going through. I have become more empathetic towards them.
Made me much more aware of his problem and to question how much of his earlier behaviour (he is now 48) may have been a result of his mental condition.
A third and minor theme was that the course refreshed or reinforced existing knowledge, skills and attitudes.
It was really a "filling in" and affirmation, and "refresher" of all I had learned and experienced in the past when working with people with cerebral palsy, head injuries, and autism. I learned a little more about other mental health problems.
Again, the responses showed that people did the course with a variety of expectations and needs, yet despite this variety, it succeeded across the board in meeting those varied needs and expectations.
Is there anything else you would like to say about the MHFA course and its value for you?
This question generated a particularly rich vein of data. The central trend that informed the vast bulk of comments was an extremely positive view of the program. Indeed, only one or two negative comments were received of any kind. Nonetheless, some themes could be discerned. First, within those who were directly positive, some were rather non-specific, while others gave detailed examples. These included gaining knowledge of how to help, greater confidence in providing help, breaking down barriers to dealing with mental health problems, and making contact with other people with similar concerns about mental health issues.
This was some of the best information I've had for years and wish I had done a course this good 30 years ago.
I found the course quite confronting. The insight gained on my personal situation was appreciated. The value of the course to me was high. I would encourage everyone to attend. I think the handbook is an excellent tool. I learnt a lot from the course.
Extremely valuable from point of view of year Advisor at a large rural high school. Provided information and avenues where to get help or refer young people to. Also valuable as a legal studies teacher (an added benefit).
Another group were very enthusiastic about the course and wanted to see it extended, either by follow on courses or by linking to new audiences.
I think it should run regularly as the population needs to learn how to deal with the situations and it creates an awareness and perhaps takes away the fear of the unknown.
This course was exceptional in its presentation and content. A more in-depth and extensive follow up program would be beneficial and valuable.
Finally, another group offered criticism. This was almost always constructive and linked to positive evaluation.
I found it very interesting, but would have liked a little more in-depth treatment of some conditions, with some various "case histories." That would have been very helpful. There seems to be a bit of a gap between how community and mental health departments are portrayed, and what they actually deliver, or take responsibility for.
Discussion
The data strongly indicate the positive value of the MHFA course. It is true that this was a non-random, availability sample and that such samples can always be challenged on the basis that they lack representation and reflect the views only of a self-selected sample, in this case perhaps of the more satisfied clients. However, if there really were some dissatisfied clients in the population, it seems very likely that at least a few would have wanted to put a critical point of view, but none were found in this study.
Moreover, the findings of this study fit with earlier quantitative evaluations which showed that the course was well received and this allows one to be even more confident of the utility of the data [10,14,15].
The qualitative data allow some more specific points to emerge beyond basic satisfaction and at least five key points should be made. First, the majority of respondents had had some direct experience, post-course, of a situation where mental health issues were salient and it is clear that for this group the course enabled them to take steps that led to better effects than otherwise might have been the case.
Second, it is clear that positive effects were experienced both intra-personally (increased empathy and understanding, increased confidence to act appropriately) as well as inter-personally (increased capacity to handle crises, manage strained relationships, offer help in an effective way, etc.)
Third, and perhaps most strikingly, the positive effects were experienced by a wide range of people including: professionals attending the course to extend their competence; family members/friends attending the course to deal with a pressing problem in their network; and individuals who themselves had experienced mental health problems. These and other groups had varied expectations and needs, yet the course was able to meet the needs of all of these groups with high levels of effectiveness.
Fourth, there is an important absence of certain kinds of potentially bad news. Any well-received course dealing with a sensitive topic runs the risk of being almost too successful. The danger would be that, flushed with enthusiasm and over-confidence, neophytes may rush in where angels fear to tread, not appreciating the dangers of their actions. If the MHFA course has had that effect, there is certainly not a trace of it in these data. This is an important and positive absence. A possible reservation on this conclusion is that those who were over-confident might not have necessarily recognized that this was the case.
Finally, those who attended were able to identify quite specific benefits and many thought the course not only immensely useful, but were keen to see it repeated and extended.
Conclusion
Examination of the post-course experiences of participants confirms the success of the MHFA course in encouraging supportive actions towards people with mental health problems. These results confirm the earlier quantitative findings from controlled trials.
Competing interests
BAK and AFJ were developers of the Mental Health First Aid course.
Authors' contributions
AFJ secured funding, was involved in the design of the study and co-wrote the paper.
BAK was involved in the design of the study and its management.
SKM had a major role in devising the questionnaire, analyzed the responses and co-wrote the paper.
All authors read and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
This study was supported by a National Health and Medical Research Council. Program Grant. We thank Kelly Blewitt and Nicole Burgess for their assistance with the study and our colleagues who were involved in the original cluster randomized trial, particularly Richard O'Kearney and Karen Peterson. The original trial was funded by the Health Promotion Demonstration Research Grants Scheme of the New South Wales Department of Health.
==== Refs
The WHO World Mental Health Survey Consortium Prevalence, severity, and unmet need for treatment of mental disorders in the World Health Organization world mental health surveys JAMA 2004 291 2581 2590 15173149 10.1001/jama.291.21.2581
Judd LL Schettler PJ Akiskal HS The prevalence, clinical relevance, and public health significance of sub-threshold depressions Psychiatr Clin North Am 2002 25 685 698 12462855 10.1016/S0193-953X(02)00026-6
Wang PS Lane M Olfson M Pincus HA Wells KB Kessler RC Twelve-month use of mental health services in the United States: results from the National Comorbidity Survey Replication Arch Gen Psychiatry 2005 62 629 640 15939840 10.1001/archpsyc.62.6.629
Caldwell TM Jorm AF Knox S Braddock D Dear KB Britt H General practice encounters for psychological problems in rural, remote and metropolitan areas in Australia Aust N Z J Psychiatry 2004 38 774 780 15369535 10.1111/j.1440-1614.2004.01461.x
Christensen H Hocking BM Smith D Web and telecounselling in Australia Med J Aust 2004 180 604 605 15200353
Getz WL Fujita BN Allen D The use of paraprofessionals in crisis intervention: evaluation of an innovative program Am J Community Psychol 1975 3 135 144 1167241 10.1007/BF00877788
Toumbourou JW Gregg ME Impact of an empowerment-based parent education program on the reduction of youth suicide risk factors J Adolesc Health 2002 31 277 285 12225740 10.1016/S1054-139X(02)00384-1
Fuller J Edwards J Martinez L Edwards B Reid K Collaboration and local networks for rural and remote primary mental healthcare in South Australia Health Soc Care Community 2004 12 75 84 14675367 10.1111/j.1365-2524.2004.00470.x
Kitchener BA Jorm AF Mental Health First Aid Manual 2002 Canberra, Centre for Mental Health Research
Kitchener BA Jorm AF Mental health first aid training for the public: evaluation of effects on knowledge, attitudes and helping behavior BMC Psychiatry 2002 2 10 12359045 10.1186/1471-244X-2-10
Albert M Becker T McCrone P Thornicroft G Social networks and mental health service utilisation – a literature review Int J Soc Psychiatry 1998 44 248 266 10459509
Kawachi I Berkman LF Social ties and mental health J Urban Health 2001 78 458 467 11564849
Jorm AF Blewitt KA Griffiths KM Kitchener BA Parslow RA Mental health first aid responses of the public: results from an Australian national survey BMC Psychiatry 2005 5 9 15694008 10.1186/1471-244X-5-9
Kitchener BA Jorm AF Mental health first aid training in a workplace setting: A randomized controlled trial [ISRCTN13249129] BMC Psychiatry 2004 4 23 15310395 10.1186/1471-244X-4-23
Jorm AF Kitchener BA O'Kearney R Dear KBG Mental health first aid training of the public in a rural area: a cluster randomized trial [ISRCTN53887541] BMC Psychiatry 2004 4 33 15500695 10.1186/1471-244X-4-33
Schank R Abelson R Wyer RS Knowledge and memory: the real story Advances in Social Cognition 1995 8 Hillsdale, New Jersey: Lawrence Erlbaum 1 85
|
16280088
|
PMC1308824
|
CC BY
|
2021-01-04 16:33:02
|
no
|
BMC Psychiatry. 2005 Nov 9; 5:43
|
utf-8
|
BMC Psychiatry
| 2,005 |
10.1186/1471-244X-5-43
|
oa_comm
|
==== Front
BMC UrolBMC Urology1471-2490BioMed Central London 1471-2490-5-141630068410.1186/1471-2490-5-14Research ArticleOffice bladder distention with Electromotive Drug Administration (EMDA) is equivalent to distention under General Anesthesia (GA) Rose Amy E [email protected] Kathryn J [email protected] Christopher K [email protected] Department of Urology, Stanford University Medical School, Stanford, California, USA2005 22 11 2005 5 14 14 8 7 2005 22 11 2005 Copyright © 2005 Rose et al; licensee BioMed Central Ltd.2005Rose et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Bladder distention is commonly used in diagnosis and treatment of interstitial cystitis (IC). Traditionally performed in the operating room under general or spinal anesthesia (GA), it is expensive and associated with short term morbidity. Office bladder distention using electromotive drug administration (EMDA) has been suggested as an alternative that is well tolerated by patients. We report the first comparative findings of patients undergoing both office distention with EMDA and distention in the operating room (OR) with GA.
Methods
This retrospective chart review identified 11 patients participating in two protocols of EMDA bladder distention who also underwent bladder distention under GA either prior to or after the EMDA procedure.
Results
The median absolute difference in bladder capacity between GA and EMDA was only 25 cc; the median percent difference was 5%. Cystoscopic findings, while not prospectively compiled, appear to have been similar.
Conclusion
This study represents the first comparison between distention with EMDA versus GA and confirms the technical feasibility of performing bladder distention in an office setting. The distention capacity achieved in the office was nearly identical to that in the OR and the cystoscopic findings very similar. Further investigation into the comparative morbidity, cost, and other outcome measures is warranted to define the ultimate role of EMDA bladder distention in the clinical evaluation and care of patients with IC.
==== Body
Background
Cystoscopy with bladder distention has traditionally been regarded as the diagnostic standard for interstitial cystitis (IC). Although there is considerable debate over the true value of bladder distention in the diagnosis of IC, the National Institute of Diabetes, Digestive and Kidney diseases (NIDDK) criteria require the presence of post-distention mucosal glomerulations (also referred to as "positive cystoscopy") or a bladder ulcer in order to qualify patients for clinical trials [1]. In addition, the International Continence Society terminology document limits the use of the term "interstitial cystitis" to patients "with typical cystoscopic and histological findings" [2]. Thus, while bladder distention is being used less commonly in the United States, it continues to be a central part of the diagnostic algorithm in most of the rest of the world [3]. Bladder distention can also provide symptomatic relief for some IC patients and is thus a commonly performed procedure for patients presenting with urinary frequency, urgency and bladder pain. The procedure is typically performed in the operating room (OR) under spinal or general anesthesia (GA) and is associated with moderate short term morbidity such as pain and hematuria. Moving bladder distention into an office setting could potentially eliminate the inherent risks of anesthesia, lower the cost of the procedure, and minimize the recovery period. In turn, this could make a complete evaluation including bladder distention available to a broader spectrum of symptomatic patients and make re-treatment practical for the subset of patients who respond favorably.
Our research group has focused on trying to develop office bladder distention as a realistic alternative to bladder distention under GA. An initial trial compared two different in-office anesthetic strategies: pre-operative intravesical alkalized lidocaine as described by Henry and colleagues [4] and lidocaine administered via electromotive drug administration (EMDA). We found that simple alkalinized lidocaine was completely inadequate for office bladder distention but that EMDA presented a promising technology. However, while we were able to perform a technically adequate bladder distention, some of our results [5] were quite different from prior studies published by Rosamilia and colleagues [6] and by Riedl [7]. We found that the majority of subjects (64%) anesthetized with EMDA were able to tolerate a 60 cm H2O distention for the full 7 minutes, and we were able to obtain a median percent increase in distention capacity over cystometric capacity of 135%. However, in contrast to prior reports that described office distention with EMDA as well tolerated, our subjects reported a median pain score at end distention of 9/10. We concluded that EMDA anesthesia was satisfactory in most cases for diagnostic in-office distention but pain was still an issue and true comparability between an EMDA distention and a distention under GA was not established.
No data exists to establish that a distention in the office with EMDA is actually equivalent to the procedure performed in the OR. We present the first report of a direct comparison between distention with EMDA versus distention with GA in a group of patients undergoing both procedures.
Methods
Two prospective protocols have been conducted to investigate the utility of EMDA anesthesia for office bladder distention. Both were approved by the Institutional Review Board. The first examined the role of EMDA distention in the initial diagnosis of IC; two patients from this protocol later went on to have a bladder distention under GA. The second protocol examined the efficacy of EMDA distention in treating patients who had previously responded to a distention in the operating room with GA. Nine subjects have enrolled in this study. The current study population is composed of these 11 patients who have experienced bladder distention in both settings.
All research carried out on human subjects was in compliance with the Helsinki Declaration. The research protocol and consent forms were approved by the Institutional Review Board (IRB) for Stanford University Medical School and informed consent was obtained from all subjects. EMDA distention was performed as an office procedure without intravenous sedation. Most patients were given oral medications for pain, anxiety, and antibiotic prophylaxis (hydrocodone 1–2 tablets, lorazepam 1 mg, ciprofloxacin 500 mg) one hour prior to the procedure. EMDA was performed as previously described [5]. Cystoscopy and distention were performed using a 15 French flexible cystoscope. The bladder was distended at 60 cm H2O under direct vision with a goal of 7 minutes distention time as measured from the start of water flow and then drained under direct vision. Pressures above 60 cm H2O for the EMDA procedure have not been attempted as not all patients could tolerate the full 7 minute distention. Pain scores were recorded during distention using a 0–10 Likert Scale. For the seven patients with Hunner's ulcers, serum lidocaine levels were obtained.
Most of the bladder distentions under GA were performed by the senior investigator (CKP) using a standard technique. Diagnostic cystoscopy was performed with a 22 Fr rigid cystoscope after which the bladder was distended at 80 cm H2O for 7 minutes under direct vision then drained under direct vision. The anesthetic technique (spinal or general) was chosen by the patient and anesthesiologist. When the prior bladder distention was performed by an outside urologist the study data was obtained from the operative note.
The comparative data was collected from the most recent distention with GA and the first EMDA distention for those patients in the treatment protocol. The patients from the diagnostic protocol had only one procedure with each technique. The date of the procedure, the distention technique, pre- and post-distention findings, and bladder capacity were recorded.
Results
Patient demographics and results are displayed in Table 1. The median age was 52 years (range 22–72); there were 6 women and 5 men. Most of the patients recruited for the treatment study had already had at least 3 previous IC therapies and many were considered end stage patients with ulcers and low bladder capacities under anesthesia. The time elapsed between the GA distention and the EMDA distention varied widely between patients with a median time of 10 months between the two procedures (range 1–91).
Table 1 Patient Demographics and Distention Results
Patient Age Sex Ethnicity Hunner's Ulcer General Anesthesia Capacity (cc)
80 cm H2O EMDA Office Capacity (cc)
60 cm H2O Absolute Difference Between OR and EMDA Capacity (cc) % Difference Pain Score During Office Distention (10)
1 61 F White + 300 260 40 13 7
2 52 F Hispanic + 275 255 20 7 9
3 68 F White + 200 205 5 2 10
4 49 F White - 900 1105 205 19 8
5 55 M White - 400 350 50 12 7
6 50 F White - 725 725 0 0 9
7* 72 M Asian + 290 370 80 22 8
8* 22 M African American - 875 850 25 3 8
9 48 F Indian + 580 550 30 5 3
10 51 M White + 400 400 0 0 8
11 59 M White + 310 300 10 3 8
Median 52 - - - - - 25 5 8
*Office distention performed prior to operating room distention. Of 11 IC patients experiencing distention both in the OR and in the office with EMDA, the distention capacity obtained in the office at 60 cm H2O very closely approximates that achieved with general anesthesia at 80 cm H2O. In patients #3,4, and 7, the capacity achieved in the office was higher than that achieved in the OR. In patients #6 and 10, the capacity achieved was exactly the same.
The bladder capacities achieved with the two different techniques were strikingly similar; the median absolute difference was only 25 cc and the median percent difference was 5%. In no case was the EMDA capacity less than 87% of the capacity achieved with GA. In three patients, we were able to achieve a higher capacity with EMDA, and in two patients, the capacities were exactly the same.
Serum lidocaine levels were drawn from the seven patients with bladder ulcers; all were less than 1.1 μg/mL. The median pain score during the distention with EMDA was 8 on a 10 point scale (range 3–10). In most patients, pain resolved rapidly after draining the bladder at the end of distention. No patient required any parenteral medication or any additional intravesical therapy for pain. Because the GA distention data was collected retrospectively we can not make comparisons about the patient experiences during the post-operative recovery.
Discussion
This is the first published comparison of bladder distention with EMDA versus GA. The data demonstrate that EMDA provides an equivalent degree of distention to the standard procedure performed in the operating room as essentially the same bladder capacity is achieved. These results confirm and enhance prior reports suggesting that office distention with EMDA is a viable alternative for select IC patients. Moving bladder distention into an office setting eliminates the risks of general anesthesia, almost certainly significantly reduces cost, and makes complete evaluation available to a wider spectrum of symptomatic patients. This is only a small retrospective series; a large scale prospective randomized controlled trial would be required to fully investigate all of the relevant factors including diagnostic utility, cost, morbidity, and therapeutic effect.
Although the distention capacities achieved in the office were almost identical to those achieved with GA, pain at the end of distention remains a problem. Patients were given low dose pre-emptive analgesia, but inevitably the pain became intense during the last one or two minutes of the distention. For most patients, however, the pain was transient and they were able to leave the office without assistance within an hour and resume full normal activities rapidly. Our pretreatment analgesic protocol was relatively conservative and could be increased without moving to conscious sedation. Most patients who have experienced both procedures prefer the transient pain of distention with EMDA to the pain and typically longer recovery associated with distention under GA.
There are important limitations to the current study. First, our study population is small and is not representative of the overall IC population. The patients are self selected for willingness to participate in an office protocol in order to avoid general anesthesia. This might lead one to believe there would be a selection bias toward patients with milder disease, but in fact most of the patients in this study are relatively refractory patients who failed standard therapies. Several of the patients would be considered "end stage" patients with severe ulcer disease and diminishing bladder capacity. The severity of disease in this patient population, however, emphasizes the potential value of EMDA anesthesia. For example, patient number 3 in Table 1 has multiple ulcers and a bladder capacity of only 200 cc at 80 cm H2O under general anesthesia, but with EMDA we were able to achieve a distention capacity of 205 cc at only 60 cm H2O with the patient fully awake. Second, it should also be noted that the time between the distention under GA and the distention with EMDA is quite variable among patients. Since most patients had EMDA second, if interval treatment improved the disease process that could bias the EMDA capacities to be higher. However, interval progression of disease could cause the opposite effect. Finally, showing identical bladder capacity with the two techniques does not prove equivalence in diagnostic or therapeutic utility. We could not report therapeutic results in this retrospective study. While we felt the cystoscopic findings are similar, a prospective study with a blinded third party analyzing the images would be required to prove this.
Conclusion
This is the first reported direct comparison between office bladder distention with EMDA versus traditional distention with general anesthesia performed in the operating room. In this select group, EMDA provided an equivalent degree of distention and thus deserves further investigation to define its ultimate clinical role. Although there is a rapid recovery, pain levels during treatment remain high in our hands. A randomized trial would be required to fully explore all of the relevant factors in comparing EMDA to GA for bladder distention.
List of abbreviations
IC: interstitial cystitis
NIDDK: National Institute of Diabetes, Digestive and Kidney diseases
OR : operating room
GA: general anesthesia
Declaration of competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Amy Rose: Made substantial contributions to conception and design, acquisition of data, analysis and interpretation of data. Was the primary person involved in drafting the manuscript and revising it critically for important intellectual content. Has given final approval of the version to be published. Involved in patient recruitment, performed the EMDA treatments, and was the primary author of the paper.
Kathryn Azevedo: Made substantial contributions to conception and design, acquisition of data, analysis and interpretation of data. Was involved in drafting the manuscript and revising it critically for important intellectual content. Has given final approval of the version to be published. Assisted in EMDA treatment and was in charge of all communication with the IRB and maintaining appropriate consent and documentation for the study.
Christopher Payne: Made substantial contributions to conception and design, acquisition of data, analysis and interpretation of data. Was involved in drafting the manuscript and revising it critically for important intellectual content. Has given final approval of the version to be published. Primary study investigator, supervised all EMDA treatments, performed OR distentions.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
Catheters, generators, and other supplies provided by the Physion© Corporation. Funding for Amy Rose provided by the "Medical Scholars" program at Stanford Medical School.
==== Refs
Gillenwater JY Wein AJ Summary of the National Institute of Arthritis, Diabetes, Digestive and Kidney Diseases Workshop on Interstitial Cystitis, National Institutes of Health, Bethesda, Maryland, August 28–29, 1987 J Urol 1988 140 203 6 3379688
Abrams P Cardozo L Fall M Griffiths D Rosier P Ulmsten U Van Kerrebroeck P Victor A Wein A The standardisation of terminology in lower urinary tract function: report from the standardisation sub-committee of the International Continence Society Neurourol Urodyn 2002 21 167 78 11857671 10.1002/nau.10052
Payne CK Terai A Komatsu K Research criteria versus clinical criteria for interstitial cystitis Int J Urol 2003 10 S7 S10 14641406 10.1046/j.1442-2042.10.s1.3.x
Henry R Patterson L Avery N Tanzola R Tod D Hunter D Nickel JC Morales A Absorption of alkalized intravesical lidocaine in normal and inflamed bladders: a simple method for improving bladder anesthesia J Urol 2001 165 1900 3 11371877 10.1097/00005392-200106000-00014
Rose AE Payne CK Azevedo K Pilot study of the feasibility of in-office bladder distention using electromotive drug administration (EMDA) Neurourol Urodyn 2005 24 254 60 15747341 10.1002/nau.20106
Rosamilia A Dwyer PL Gibson J Electromotive drug administration of lidocaine and dexamethasone followed by cystodistention in women with interstitial cystitis Int Urogynecol J Pelvic Floor Dysfunct 1997 8 142 5 9449586
Riedl CR Knoll M Plas E Pfluger H Electromotive drug administration and hydrodistention for the treatment of interstitial cystitis J Endourol 1998 12 269 72 9658301
|
16300684
|
PMC1308825
|
CC BY
|
2021-01-04 16:30:02
|
no
|
BMC Urol. 2005 Nov 22; 5:14
|
utf-8
|
BMC Urol
| 2,005 |
10.1186/1471-2490-5-14
|
oa_comm
|
==== Front
BMC Vet ResBMC Veterinary Research1746-6148BioMed Central London 1746-6148-1-91628197110.1186/1746-6148-1-9Research ArticleEquine herpesvirus 2 (EHV-2) infection in thoroughbred horses in Argentina Craig María I [email protected] María E [email protected]ández Fernando M [email protected] Instituto de Virología, Centro de Investigación en Ciencias Veterinarias y Agronómicas (CICVyA), INTA, CC 25, (1712) Castelar, Buenos Aires, Argentina2005 9 11 2005 1 9 9 15 6 2005 9 11 2005 Copyright © 2005 Craig et al; licensee BioMed Central Ltd.2005Craig et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Equine herpesvirus 2 is a gamma-herpesvirus that infects horses worldwide. Although EHV-2 has been implicated in immunosuppression in foals, upper respiratory tract disease, conjunctivitis, general malaise and poor performance, its precise role as a pathogen remains uncertain. The purpose of the present study was to analyse the incidence of EHV-2 in an Argentinean horse population and correlate it with age and clinical status of the animals.
Results
A serological study on 153 thoroughbred racing horses confirmed the presence of EHV-2 in the Argentinean equine population. A virus neutralization test showed a total of 79.7 % animals were sero-positive for EHV-2. An increase in antibodies titre with age as well as infection at earlier ages were observed.
EHV-2 was isolated from 2 out of 22 nasal swabs from horses showing respiratory symptoms. The virus grew slowly and showed characteristic cytopathic effect after several blind passages on RK13 cells. The identity of the isolates was confirmed by nested PCR and restriction enzyme assay (REA).
Conclusion
This is the first report on the presence of EHV-2 in Argentina and adds new data to the virus distribution map. Though EHV-2 was isolated from foals showing respiratory symptoms, further studies are needed to unequivocally associate this virus with clinical symptoms.
==== Body
Background
Equine herpesvirus 2 (EHV-2) is a slowly growing, cell-associated gamma-herpesvirus. This virus is widespread throughout the equine population and has been isolated from horses of different countries like United Kingdom [1], Japan [2], Australia [3], New Zealand [4], Switzerland [5], Germany [6,7], United States [8,9], Canada [10], Hungary [11] and more recently, from Poland [12]. Although its role as a pathogen is controversial, some authors have reported its association with upper respiratory tract disease, inappetance, lymphadenopathy, immunosuppression, keratoconjunctivitis, general malaise and poor performance [8,13,7,16].
Equine Influenza Virus and Equine Herpesvirus 4 (EHV-4) are the most common viral agents related to respiratory disease in Argentina (Dr. Barrandeguy, personal communication). These and other respiratory viruses as Adenovirus, Equine herpesvirus 1 (EHV-1), Arteritis Virus and Rhinovirus, are usually checked in the diagnostic routine of our laboratory but no records about EHV-2 isolation or frequency of sero-positive samples were available.
The purpose of the present study was to analyse the presence of EHV-2 in an Argentinean horse population and to correlate its incidence with age and presence of respiratory symptoms.
Results and discussion
Sero-prevalence of EHV-2 was calculated on one hundred and fifty-three (153) thoroughbred racing horses by a neutralization test (NT). Cross reactivity with EHV-5 was not checked in this study. The percentage of sero-prevalence to EHV-2 was 79.7% (122/153). Sera samples were grouped according to the clinical status into animals with symptoms (fever, cough, nasal discharge) or clinically healthy. Again, each of these groups was divided according to the animal age in older or younger than 1 year old. The arithmetical mean of the antibody titres was calculated for each of these four groups. Mean antibody titres between older and younger than 1 year old animals both, with and without clinical symptoms were statistically compared. Mean values for the older horses (1.28 > 1.02) were significantly (p < 0.05) higher than for the younger ones. These results agree with the observations of other authors [17] about the increase in antibodies titre with age. Within the older than 1-year group, the mean titre in the group with clinical symptoms was higher (1.34>1.22) though not significant. However, mean titre values were significantly (p < 0.05) higher (1.19>0.85) in the group with clinical symptoms within the younger than 1 year group (Table 1). This difference might be related to the early exposure to this agent.
Table 1 Distribution of serum samples according to age and clinical status
With symptoms Antibodies Clinically healthy Antibodies
Animals Sero-negative Sero-positive (Mean titre)a Sero-negative Sero-positive (Mean titre)a
Older than 1 year old 5 32 1.34 3 40 1.22
Younger than 1 year old 3 26 1.19 18 26 0.85*
Total (n) 66 87
n: number of samples
a: arithmetical media of Ab titre (Reed and Muench)
*: Significantly different (p < 0.05)
Our results suggest the virus is circulating with a high prevalence on the analysed equine population, in accordance with other sero-prevalence data [17,18], and confirm previous reports [19] about the acquisition of EHV-2 at earlier ages.
Taking into account the relatively high percentage of sero-prevalence and the association of EHV-2 with respiratory disease [16], the isolation of this viral agent from horses showing different respiratory symptomathology was carries out. Twenty-two (22) nasal swabs from horses, aged between 6 months and 2 years old, displaying respiratory symptoms were checked for the respiratory viruses commonly analysed in the laboratory routine, and none of them resulted positive for these viruses. Only two (2) nasal swab samples, named E1 and E2, showed CPE after the third blind passage. Some authors reported the presence of vacuoles in RK-13 cells infected with EHV-2 isolates [20] while others described various CPE forms depending on the virus isolate and cell type [21]. In our isolate, CPE was characterized by rounded cells, syncytia and vacuolized cell aggregates with suited partial cell membrane fusion. The late development of CPE referred to a slow growing viral agent.
In order to identify these two isolates, two simple techniques like EM and IIFA were first used. The particle morphology observed by EM showed a size corresponding to a typical herpesvirus (data not shown). The IIFA showed fluorescence signals on cells infected with EHV-2 reference strain LK, E1 and E2. Mock-infected cells were negative (data not shown). As the polyclonal antiserum used for the staining could have cross reaction with EHV-5, another member of the gamma-herpesvirinae subfamily that shows growth characteristics similar to EHV-2 [22], a virus specific nested PCR and a restriction enzyme analysis (REA) were performed.
Nested PCR was carried out on E1 and E2 samples as described in Methods. After the second round, both isolations showed a band of around 600 bp (Figure 1), as the one displayed by the positive control.
Figure 1 Nested PCR on nasal swab samples from foals with respiratory symptoms. Nested PCR amplification products from two equines (E1 and E2) showing respiratory symptoms were analysed by gel electrophoresis. DNA was extracted from original nasal swabs; DNA from the suspension was used as template in the first round. Genomic DNA from EHV-2 strain LK was used as a PCR positive control (PCR+) and H20 as negative control (PCR -). DNA marker φX-174 Hae and λ-RF/Hind III is seen in lane M.
EHV-5 identity of the isolations was unequivocally excluded by REA. As shown in Figure 2, the patterns after digestion with Hind III and Eco RI of the isolates and a control EHV-5 were completely different.
Figure 2 Restriction enzyme assay profile of EHV-2 isolates and reference strains. EHV-2 isolates (E1 and E2), EHV-2 strains H40, T400, and LK4, and EHV-5 strain P48, were amplified in RK-13 cell monolayer. Viral DNA was purified by Genomic-tip 20/G column, digested with Hind III and Eco RI and analysed by gel electrophoresis. DNA marker φX-174 Hae III and λ-RF/Hind III is seen in lane M.
As it was specified by others authors [23,24], EHV-2 was described as a virus with a high genomic variation. In our experiments, the restriction patterns of the isolates were different to those of T400 and LK4. Moreover, E1 and E2 resulted identical to each other and very similar to strain H40. Both, E1 and E2 belonged to the same stud farm and were probably epidemiologically related.
Conclusion
Although EHV-2 has been reported in equine populations worldwide, no reports about its prevalence in Argentina were available. In this study, serological data and virus isolation provided a clear evidence for the presence of this agent on the studied horse population. In horses younger than 1 year old but not in the older ones, serological data correlated with the presence of clinical symptoms. Though viral agents associated with respiratory disease were discarded, no bacterial analysis was done in the studied samples. Further studies are necessary to correlate EHV-2 with the respiratory disease. The EHV-2 identity of the isolated viruses was finally confirmed by the use of a nested PCR and a restriction enzyme assay. This is the first report of EHV-2 isolation in Argentina and adds new data to the EHV-2 distribution map.
Methods
Cells
RK-13 (ATTCC CCL 37) cell monolayers were grown with Eagle 's- minimum essential medium (MEM-E) supplemented with 100 UI/ml penicillin, 60 μg/ml streptomycin, 50 μg/ml gentamycin and 10% FCS. Subsequently, the cells were maintained with MEM-E supplemented with same concentrations of antibiotics and 2% FCS.
Virus
EHV-2 strains LK (ATTCC VR 701), H40 (NVSL Ames Iowa), LK4 and T400 as well as, EHV-5 strain P48, were kindly supplied by Dr. Borchers (Institut für Virologie, Freie Universität, Berlin-Germany).
Samples
Sera
Serum samples were obtained from one hundred and fifty three (153) thoroughbred racing horses, aged between 6 months and 2 years old.
Nasal swabs
Nasal swabs from twenty-two horses showing respiratory symptoms like nasal discharge, cough and fever, were used in our research. Samples were collected with a cotton swab and transferred to a tube containing 3 ml of E-MEM supplemented with 2% FCS and antibiotics. Clarification was performed by centrifugation at 3000 rpm (Hermle Z 513 K) for 15 min. Each sample was aliquoted and stored at -70°C until use.
Neutralization test
This assay was performed on 96 wells RK-13 cell monolayer, where serial four-fold dilutions of complement-inactivated serum were incubated with 100 ID50 of LK strain for 4 hs at 37°C as described elsewhere [25]. After 7 incubation days, antibody titres were calculated by Reed and Muench formula [26]. Sera that protected 100% of the cell monolayer at the lower dilution assayed were considered positive. An EHV-2 positive horse reference serum used as positive control, was kindly supplied by Dr. Borchers.
Indirect immunofluorescence assay
Trypsinised virus infected cells were seeded on IFA micro slides and fixed with ice-cold acetone for 30 min. A polyclonal EHV-2 specific rabbit serum (341 EDV 8301, NVSL Ames Iowa) was added and incubated in a moist chamber al 37°C for 1 h. Fluorescein conjugated rabbit antiserum (KPL Kikergaard and Perry laboratories) was used to reveal the reaction. EHV-2 LK strain infected RK-13 and non-infected cells were used as positive and negative assay controls, respectively.
Electron microscopy
Cell cultures showing 70% CPE were trypsinised and sonicated for 30 seconds in a Sonicator XL (Hert System) and spinned at 3000 × g during 10 minutes to eliminate cell debris. The supernatant was centrifuged at 14000 × g for 1 h 30 minutes and the resulting pellet was resuspended with a saline buffer solution and stained with 2% phosphotungstate acid.
Virus isolation
All nasal swab samples were previously checked for the presence of equine viruses associated with respiratory disease. Equine adenovirus, rhinovirus, arteritis virus, equine herpesvirus 1 and 4, were checked by isolation in cell tissue culture while equine influenza virus was assayed using embryonated eggs. Once it was confirmed that all the samples were negative for the mentioned viruses, the presence of EHV-2 was analysed. Briefly, 0.5 ml of each sample was inoculated onto confluent RK-13 monolayers in 25 cm2 flask (NUNC™, Denmark). After an adsorption period at 37°C for 1 h, the inoculum was removed and cells were maintained at 37°C at an atmosphere of 5% CO2 for 7 days with periodic microscopical examination. After this time period, a blind passage was made if CPE was absent. Cell monolayers were thawed and freezed three times and an aliquot of the suspension was passed on to fresh cell culture. The samples were considered negative after the forth-blind passage.
Nested PCR
The nested PCR used in this study amplifies a 620 bp sequence located upstream of the ORF coding for vIL-10 [27].
Original nasal swab samples were centrifuged for 10 min at 15,000 × g as described elsewhere [27]. Briefly, resulting pellets were diluted in 50 μl digestion buffer (50 mM Tris pH 8.5; 1 mM EDTA; 0.5% Tween 20) plus 1.5 μl Proteinase K (10 mg/ml, Sigma) and incubated at 56°C for 3 h, boiled for 10 min at 95°C to inactivate the proteinase K and the supernatant used for nPCR. PCR conditions were optimised using purified viral DNA from the EHV-2 LK strain obtained as described below. The PCR mix (50 μl) contained 0.2 mM dNTPs, 0.4 μM of each primer [27], 1.5 U Taq polymerase (QIAGEN), 1 × reaction buffer (QIAGEN) and 10 μl of DNA suspension obtained from the original nasal swab sample. The optimal annealing temperature for each of the two primer pairs was 60°C. Cycling was carried out using a Biometra trio-thermblock (Biometra®) with 35 amplification cycles consisting of denaturation at 94°C for 30 sec, annealing at 60°C for 30 sec and extension at 72°C for 1 min. For the second round of this nested PCR, 3 μl from the first PCR reaction were amplified with the inner pair of primers using the same cycle as before. Purified viral DNA from the EHV-2 LK strain was used as a PCR positive control.
Restriction enzyme assay
Viral DNA was obtained from RK-13 cells infected with EHV-2 isolates and also with EHV-2 reference strains H40, LK4, T400 and EHV-5 (P48). Viruses were pelleted from cell culture supernatant by centrifugation through a 30% sucrose cushion at 24,000 × g for 1 h 30 min. Pellets were resuspended with DNA buffer (10 mM MgCl2, 2 mM CaCl2, 200 mM Tris pH 7.4) and treated with RNAse (10 mg/ml) and DNAse I (2 mg/360 μl) at 37°C for 1 h. Subsequently, the pellets were treated with EDTA (0.2 M), 1% sodium dodecyl sulphate and proteinase K (10 mg/ml) (Boehringer, Manheim). Finally, DNA was purified by Genomic-tip 20/G column (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. Aliquots of 200 ng of the obtained purified viral DNA were digested overnight at 37°C with Eco RI and Hind III. The fragments were separated by electrophoresis through 1% agarose gel and stained with ethidium bromide.
Statistical analysis
Two (2) ways ANOVA [28] followed by post hoc comparisons (Tukey HSD for unequal N, Spiotvoll/Stoline test) [29] were used.
List of abbreviations
EHV-2, equine herpesvirus 2; EHV-5, equine herpesvirus 5; E-MEM, Eagle's minimum escential medium; EM, electron microscopy; IIFA indirect immunofluorescence assay; Ab, antibody; PCR, polymerase chain reaction; REA, restriction enzyme assay; NT, neutralization test; CPE, cytopathic effect; vIL-10, viral interleukin 10; ORF, open reading frame.
Authors' contributions
M.I. Craig carried out the experimental work and drafting of the manuscript. M.E. Barrandeguy carried out the data collection procedure and conceived the study. F.M. Fernández coordinated the project. All authors read and approved the final manuscript.
Acknowledgements
This work was supported by the Technology Link Agreement INTA-Haras and a CONICET fellowship.
The authors thank Dr. Kerstin Borchers for useful comments and critical review of this manuscript. Technical assistance of O. Zabal, D. Compaired and T. Leiskau is gratefully acknowledged. We thank Dr. A. Bolondi for his assistance in electron microscopy.
==== Refs
Plummer G Waterson AP Equine herpesvirus Virol 1963 19 412 416 10.1016/0042-6822(63)90083-7
Kono Y Kobayashi K Cytopathogenic equine orphan (CEO) virus in horse kidney cell culture I. Isolation and properties Natl Inst An Health Q Japan 1964 4 10 20
Studdert MJ Turner AJ Peterson JE Equine herpesvirus 1. Isolation and characterization of equine rhinopneumonitis virus and other equine herpesvirus from horses Aust Vet J 1970 46 83 89 4317854
Horner GW Hunter R O'Flaherty JD Dickinson LG Isolation of equine herpesvirus from horses with respiratory disease N Z Vet J 1976 24 171 176 1070615
Karpas A Characterization of a new herpes-like virus isolated from foal kidney Ann Inst Pasteur, Paris 1966 110 688 696 5909615
Petzoldt K Die Virusausscheidg beim Stutenabort (Rhinopneumonitis) Arch Exp Vetarmed 1967 21 115 119
Thein P Bryans JT, Gerber H The association of EHV-2 infection with keratitis and research on the occurrence of equine coital exanthema (EHV-3) of horses in Germany Equine infectious disease IV 1978 Princeton New Jersey: Veterinary Publications 33 41
Kemeny LJ Pearson JE Isolation of herpesvirus from equine leucocytes: comparison with equine rhinopneumonitis virus Can J Comp Med 1970 34 59 65 4246005
Roberts AW Whitenack DL Carter GR Recovery of adenoviruses and slow herpesviruses from horses having respiratory tract infection Am J Ve t Res 1974 35 1169 1172
Sherman J Thorsen J Barnum JA Mitchell WR Ingram DG Infectious causes of equine respiratory disease on Ontario Standardbred racetracks J Clin Microbiol 1977 5 285 289 192757
Pàlfi V Belák S Molnár T Isolation of equine herpesvirus type 2 from foals showing respiratory symptoms Zentralbl Veterinarmed B 1978 25 165 167 207057
Rusczcyk A Cywinska A Banbura MW Equine herpes virus 2 infection in horse populations in Poland Acta Virol 2004 48 189 192 15595214
Plummer G Goodheart CR Studdert MJ Equine herpesviruses: antigenic relationships and deoxyribonucleic acid densities Infect Immu 1973 8 621 627
Studdert MJ Comparative aspects of equine herpesviruses Cornell Vet 1974 64 94 122 4359988
Blakeslee JR Olsen RG Mc Allister ES Fassbender J Dennis R Evidence of respiratory tract infection induced by equine herpesvirus type 2 in the horse Can J Microbiol 1975 21 1940 1946 175904
Belák S Pàlfi V Tuboly S Bartha L Passive immunization of foals to prevent respiratory disease caused by equine herpesvirus type 2 Zentralbl Veterinarmed B 1980 27 826 830 6164185
Bagust TJ Pascoe RR Harden TJ Studies on equine herpesviruses 3. The incidence in Queensland of three different equine herpesvirus infections Aust Vet J 1972 48 47 53 4335284
Mc Guire TC Crawford TB Henson JB Prevalence of antibodies to herpesvirus types 1 and 2, arteritis and infectious anemia viral antigens in equine serum Am J Vet Res 1974 35 181 185 4360338
Harden TJ Bagust TJ Pascoe RR Spradbow PB Studies on equine herpesvirus 5. Isolation and characterization of slowly cytopathic equine herpesviruses in Queensland Austr Vet J 1974 50 483 488
Roeder PL Scott GR The prevalence of equid herpes virus 2 infections The Vet Rec 1975 96 404 405 167501
Browning GF Studdert MJ Equine herpesvirus 2 (Equine Cytomegalovirus) Vet Bull 1998 58 775 790
Agius CT Studdert MJ Equine herpesviruses 2 and 5: comparisons with other members of subfamily gamma-herpesvirinae Adv Virus Res 1994 44 357 379 7817877
Browning GF Studdert MJ Physical mapping of a genome of equine herpesvirus type 2 (equine cytomegalovirus) Arch Virol 1989 104 77 86 2923549 10.1007/BF01313809
Collinson PN O'Reilly JL Ficorilli N Studdert MJ Isolations of equine herpesvirus type 2 (equine gammaherpesvirus 2) from foals with keratoconjunctivitis J Am Vet Med Assoc 1994 205 329 331 7928614
Nordenghran A Kingeborn B Lindholm A Merza M The use of neutralizing monoclonal antibody to detect infections of equine herpesvirus type 2 (EHV-2) J Vet Diagn Invest 2001 13 389 393 11580059
Reed L Muench H A simple method of estimating fifty per cent end points Am J Hyg 1938 27 493
Borchers K Wolfinger U Goltz M Broll H Ludwig H Distribution and relevance of equine herpesvirus type 2 (EHV-2) infections Arch Virol 1997 142 917 928 9191857 10.1007/s007050050128
Sokal RR Rohlf FJ Biometry 1981 2 Freeman, New York 1 859
Winer BJ Statistical Principles in Experimental Design 1971 2 Mac Graw Hill, New York 907
|
16281971
|
PMC1308826
|
CC BY
|
2021-01-04 16:29:46
|
no
|
BMC Vet Res. 2005 Nov 9; 1:9
|
utf-8
|
BMC Vet Res
| 2,005 |
10.1186/1746-6148-1-9
|
oa_comm
|
==== Front
J CarcinogJournal of Carcinogenesis1477-3163BioMed Central London 1477-3163-4-201627765810.1186/1477-3163-4-20ResearchRole of retinoic acid receptors in squamous-cell carcinoma in human esophagus Bergheim I [email protected] E [email protected] E [email protected]ölscher AH [email protected] Ch [email protected] A [email protected] Department of Physiology of Nutrition, Hohenheim University (140e), Garbenstrasse 28, 70599 Stuttgart, Germany2 Department of Visceral and Vascular Surgery, University of Cologne, Joseph-Stelzmann-Strasse 9, 50931, Cologne, Germany2005 8 11 2005 4 20 20 19 7 2005 8 11 2005 Copyright © 2005 Bergheim et al; licensee BioMed Central Ltd.2005Bergheim et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Worldwide, cancer in the esophagus ranks among the 10 most common cancers. Alterations of retinoic acid receptors (e.g. RARα, β, γ, and RXRα, β, γ) expression is considered to play an important role in development of squamous-cell carcinoma (SCC), which is the most common esophageal cancer. Alcohol consumption and smoking, which can alter retinoic acid receptor levels, have been identified as key risk factors in the development of carcinoma in the aero-digestive tract. Therefore, the aim of the present study was to evaluate protein levels of retinoic acid receptors (i.e. RARα, β, γ, and RXRβ) in esophageal SCC and surrounding normal tissue of patients with untreated SCC and controls.
Methods
All study participants completed a questionnaire concerning smoking and alcohol drinking habits as well as anthropometrical parameters. Protein levels of RARα, β, γ, and RXRβ were determined by Western Blot in normal esophageal tissue and tissue obtained from SCC of 21 patients with newly diagnosed esophageal SCC and normal esophageal tissue of 10 controls.
Results
Protein levels of RARγ were significantly lower by ~68% in SCC compared to normal surrounding tissue in patients with SCC that smoked and/or consumed elevated amounts of alcohol. Furthermore, RARα protein levels were significantly lower (~- 45%) in SCC in comparison to normal esophageal mucosa in patients with elevated alcohol intake. When comparing protein levels of retinoic acid receptors between normal tissue of patients with SCC and controls, RARγ protein levels were found to be significantly higher (~2.7-fold) in normal esophageal tissue of SCC patients than in esophageal tissue obtained from controls. No differences were found for RARα, β, and RXRβ protein levels between normal esophageal tissue of patients and that of controls.
Conclusion
In conclusion, results of the present study suggest that alterations of retinoic acid receptors protein may contribute in the development of SCC in esophagus and that in some patients life style (e.g. smoking and alcohol consumption) may be a critical component in the alteration of retinoic acid receptor levels in esophagus.
==== Body
Background
Worldwide esophageal cancer ranks among the ten most common cancers [1] and the overall 5-year survival rate for patients diagnosed with esophageal cancer is poor (e.g. 3%–10%) [2]. Although, there is a rising entity of adenocarcinoma [3], the majority of carcinoma of the esophagus are squamous-cell carcinoma (SCC) [1]. Multiple epidemiological studies indicate that alcohol and tobacco may be important risk factors for the development of carcinoma in the esophagus [1,4]. In a case-control study, Valsecchi et al. [5] showed that people who smoked 40 or more years had almost 6-times the risk of esophageal cancer development compared to non-smokers. Those with a history of alcohol consumption of more than two beers per day had 3-times the risk in comparison to abstainers. Furthermore, the combination of both habits was found to have a synergistic effect, increasing the relative risk for developing cancer in the aero-digestive tract up to 17.3-times over that of the non-smokers, non-drinkers [6]. However, despite intense research [7], molecular pathomechanisms involved in the development of esophageal SCC are not fully understood so far contributing to the lack of successful pharmacological therapies.
Alterations of retinoid metabolism and retinoic acid receptor (e.g. RARα, β, γ, and RXRα, β, γ)-mediated gene transcription have been proposed to play an important role in the pathogenesis of SCC [8-10]. It is well established that RARs and RXRs are ligand-dependent transcription factors with distinct expression patterns in early and adult stages of development. Therefore, down-regulation (e.g. loss or low expression of the specific nuclear receptor) or "functional" down-regulation (e.g. lack of ligands) of both retinoic acid receptor families could interfere with the retinoid signal transduction and, over time, might result in enhanced cell proliferation and malignant transformation. An association between altered expression of nuclear retinoic acid receptors and the malignant transformation of human cells has been demonstrated in several studies [11-14]. Furthermore, a reduced expression of RARγ found in head and neck SCC cell lines [15] seems to depend on a sufficient Vitamin A supplementation [16].
Since little is known about the abundance of RARs and RXRs protein in esophageal SCC in humans, the major focus of the present study was to determine protein levels of RARα, β, γ, and RXRβ in normal, macroscopically unaffected esophageal tissue and SCC of patients with untreated SCC and controls. In addition, alcohol intake and smoking habits of patients and controls were evaluated and related to retinoic acid receptor levels.
Materials and methods
Subjects and tissue specimens
The study was approved by the Ethics Committee of the Medical Clinic of the University of Cologne, Germany. Informed consent was obtained from all subjects included in the study. A total of 31 subjects, all of whom were undergoing endoscopies for medical reasons, were included in the study. Twenty-one patients had untreated SCC of the esophagus, and ten SCC-free subjects served as controls. Histopathological analysis of SCC was performed by an experienced physician. Tumor staging was based on a graded rating system and varied from G1 (well differentiated) to G3 (poorly differentiated). In controls, the presence or absence of SCC was confirmed macroscopically. All study participants completed a questionnaire concerning smoking and alcohol drinking habits as well as anthropometrical parameters. Using standard pinch forceps, two biopsies were obtained from macroscopically normal esophageal tissue and SCC of subjects with carcinoma and from the mucosa of healthy controls. Biopsies were placed immediately in liquid nitrogen and stored at -80°C until analysis.
Immunoblot Analysis
Total protein was isolated using Trizol reagent (Invitrogen, Carlsbad, CA, USA). Using SDS-polyacrylamide gel electrophoresis, 20–30 μg of total protein were separated in a 9% polyacrylamide gel (PAGE). Protein was transferred to a nitrocellulose membrane (Amersham/ Pharmacia, Freiburg, Germany). Membranes were blocked in Tris-buffered saline (TBS) with 5% non-fat milk powder with Tween 20 (TBST, 0.01% v/v Tween 20), rinsed in TBST, and probed with primary antibodies dissolved in TBS followed by an incubation with the secondary antibody. Primary antibodies against RARα, β, γ, and RXRβ were obtained from Alexis (Dianova, Hamburg, Germany), the secondary peroxidase-conjugated antibodies were purchased from Boehringer (Mannheim, Germany). Bands were visualized by enhanced chemiluminescence using SuperSignal® West Dura (Pierce, KTM, Bad Godesberg, Germany). Blots were photographed (Camera LAS 1000, Fuji, USA) and immunoquantitation was accomplished by densitometric analysis using the software AIDA (Raytest, Isotopenmessgeraete, Straubenhardt, Germany). To achieve standardization and comparability of blots, total protein derived from human liver was loaded on each gel at three concentrations (50 μg, 25 μg, and 12.5 μg) and blottet with the samples. To ensure equal loading, all blots were stained with Ponceau red; haptene signals were normalized to β-actin using a commercially available antibody (Sigma Chemical Co., Munich, Germany).
Statistical Analysis
Results are presented as mean ± standard error of the mean (SEM), unless otherwise stated. Wilcoxon t test was used to test for significance of differences in expression levels measured in both normal and malignant tissue of patients with SCC. Fisher's exact test was used for comparison of lifestyle data. Statistical comparison of values originating from separate groups was performed with the Mann-Whitney U-test. P < 0.05 was considered to represent a significant difference.
Results
Subject age ranged from 38–71 years and most were of normal weight (Table 1). Smoking habits did not differ between patients and controls. However, significantly more patients with SCC reported to consume "elevated" amounts of alcohol (more than 30 g of alcohol/d or more 4-times a week more than 1 drink) in comparison to controls (Table 1). Tumor grading ranged form G1 to G3 with the majority being G2 (Table 2).
Table 1 Characteristics of patients with untreated esophageal SCC and controls.
Parameter Patients Controls p-values
n 21 10
Age 56.7 ± 1.9 51.1 ± 3.5 0.118
Sex (female/male) 7/14 5/5 0.425
BMI 23.8 ± 1.4 25.1 ± 0.6 0.370
Cigarette usage (yes/no) 10/11 3/7 0.242
Alcohol consumption*)
none 4**) 2 0.012
moderate 8 8
elevated 9 0
All data expressed as mean ± SEM. (BMI = body mass index, p < 0.05 statistically significant) *) none = study subject reported to never drink alcohol; moderate = study subject reported to drink alcohol sometimes (not more than once a week); elevated = study subject reported to consume alcohol more than 4-times a week (more than one drink) or more than 30 g ethanol/d on the average **) two patients had been former heavy alcohol abusers but reported to have given up alcohol consumption several years ago
Table 2 Tumor staging of patients with SCC.
Tumor stage n
G1 2
G1-2 2
G2 8
G2-3 4
G3 5
G1 = well differentiated
G2 = moderately differentiated
G3 = poorly differentiated
Retinoic acid receptor protein levels in patients with untreated esophageal SCC
Figures 1 and 2 summarize the results of protein measurements performed in biopsy specimens obtained from SCC and neighboring normal esophageal tissue of patients with untreated esophageal SCC. Western Blot analyses demonstrated that RARα, β, γ, and RXRβ were present as a major immunoreactive band representing the 55 kDa monomer in all tissue samples tested (see Figure 1). Protein levels of RARγ were found to be significantly lower in specimens obtained from SCC in comparison to surrounding normal tissue. Specifically, RARγ protein levels were ~47% lower in SCC when compared with those of normal, unaffected esophageal tissue (p = 0.040). In contrast, no significant differences were found when comparing protein levels of RARα, RARβ, and RXRβ between macroscopically normal, unaffected esophageal tissue and SCC.
Figure 1 RARα, β, γ, and RXRβ protein in human esophagus. Representative Western blot of RARα, β, γ, and RXRβ protein in normal esophagus of controls and normal, unaffected esophageal tissue of patients with untreated esophageal SCC and SCC. Protein levels of β-actin were determined as loading control.
Figure 2 Protein levels of RARα, β, γ, and RXRβ in patients with esophageal SCC. Quantitative analysis of Western blots performed in normal tissue of patients with SCC and tissue obtained form SCC. Protein levels of β-actin were determined as loading control. Data are means ± SEM. *p < 0.05
To investigate whether smoking and alcohol consumption influences protein levels of esophageal retinoic acid receptors of patients with SCC were grouped by smoking habits (e.g. non-smoker or smoker) regardless of their alcohol consumption. In smokers, RARγ protein levels were found to be significantly lower (~68%, p = 0.010) in SCC in comparison to unaffected esophageal tissue in smoking patients (n = 10). Interestingly, a similar difference was not found in patients who were non-smokers (n = 11, data not shown). Similar, patients were than sub-grouped by alcohol intake regardless of their smoking habits into patients who (a) reported to consume no alcohol (= "none"), (b) who reported to be "moderate" alcohol consumers not exceeding alcohol intake more than once a week, or (c) who stated to consume alcohol more than 4-times a week or more than 30 g raw ethanol/ d (see also Table 2). Since only 4 patients reported to be abstainers and since two patients in this group stated to have given up heavy alcohol consumption several years ago, this group was excluded from the statistical analysis. Results are summarized in Figure 3. When comparing protein levels of retinoic acid receptors between SCC and unaffected tissue obtained from patients who reported to consume "elevated" amounts of alcohol (n = 9), RARγ protein levels were again found to be significantly lower in SCC in comparison to neighboring normal tissue (~40%, p = 0.038). Furthermore, in this subgroup of patients, in biopsies obtained from SCC protein levels of RARα were ~55% lower in comparison to normal tissue (p = 0.049). Again, similar to non-smokers, in patients who were moderate consumers of alcohol (n = 8), neither protein levels of RARα nor RARγ were found to differ between tissue obtained from SCC and normal esophageal tissue (data not shown). Next, receptor protein levels were compared between SCC and normal tissue of patients who consumed elevated amounts of alcohol and smoked (n = 6). Interestingly, only levels of RARγ were found to be significantly lower by ~68% in esophageal SCC in comparison to normal esophageal tissue (p = 0.046). As to be expected, in patients, who had a "moderate" alcohol consumption and were non-smokers, no differences were found in protein levels of RARs or RXRβ.
Figure 3 Protein levels of RARα, β, γ, and RXRβ in smoking and alcohol consuming patients with esophageal SCC. Quantitative analysis of RARα, β, γ, and RXRβ protein levels in normal unaffected esophageal tissue und SCC of (A) smoking SCC patients, (B) patients with "elevated" alcohol consumption, and (C) smoking SCC patients with "elevated" alcohol intake. Protein levels of β-actin were determined as loading control. Data are means ± SEM. *p < 0.05 (Normal tissue = macroscopically unaffected tissue, SCC = squamous-cell carcinoma)
Retinoic acid receptor protein levels in normal tissue of patients with esophageal SCC and controls
Since it has been suggested by the results of others (19) that basal levels of retinoic acid receptor mRNA expression in normal unaffected tissue might differ between patients with SCC and controls, protein levels of RARα, β, γ, and RXRβ were determined in esophageal tissue obtained form controls and compared with protein levels measured in normal esophageal tissue obtained from patients with esophageal SCC. Representative Western blots are depicted in Figure 1 A and quantitative analysis of blots is summarized in Figure 4. Analysis revealed a significantly higher protein concentration of RARγ in normal unaffected tissue of patients with esophageal SCC in comparison to controls. Specifically, protein levels were found to be ~2.7-fold higher in normal tissue of patients when compared with controls (p = 0.048). No significant differences were found when comparing protein levels of RARα, γ, and RXRβ measured in normal, macroscopically unaffected tissue specimens of patients and controls.
Figure 4 RARα, β, γ, and RXRβ protein levels in normal esophageal tissue of patients with SCC and controls. Quantitative analysis of RARα, β, γ, and RXRβ protein levels as determined by Western blot. Protein levels of β-actin were determined as loading control. Data are means ± SEM. *p < 0.05
Discussion
Alterations of RARγ and RARα protein levels in patients with esophageal SCC are depending on life style
In the present study, the hypothesis was tested that altered protein levels of retinoic acid receptors might be involved in the development of esophageal SCC in humans. Indeed, protein levels of RARγ were found to be significantly lower in SCC relative to normal tissue in patients who smoked and consumed alcohol. A similar difference was found for RARα in SCC patients who reported to consume "elevated" amounts of alcohol. However, diminished levels of RARγ and RARα were only found in these sub-populations of patients with SCC in the esophagus, suggesting that underlying mechanism leading to the development of SCC may vary depending on life style of patients. Several groups have previously determined mRNA expression of RARs and RXRs in normal esophageal tissue and SCC. Zhang et al. [17] and Qiu et al. [18], who measured mRNA levels of RARs and RXRs found a striking reduction of the expression of RARβ but not of RARγ in malignant esophageal tissue in comparison to normal tissue. Furthermore, in in situ studies a reduced mRNA expression of RARα was found in esophageal SCC compared with normal tissue [11]. However, in these studies, no information was given on life style parameters such as smoking and alcohol intake and patients were not sub-grouped accordingly and only mRNA expression levels of RARs and RXRs were evaluated.
Due to the retrospective character of the present study it can only be speculated whether the association of cigarette smoking and reduced protein levels of RARγ in SCC is a cause – and effect relationship or not. However, the action of tobacco carcinogens are believed to be mediated through covalent binding to DNA, RNA, or protein, forming DNA, RNA, and protein adducts [19]. Furthermore, results of in vitro and in vivo studies [20] suggest that tobacco carcinogens (e.g. benzo(a)pyrene diol epoxide) might lead to DNA methylation of RAR genes subsequently resulting in an inhibition of the expression of RARs rather than mediating their action through direct mutations.
Several mechanism have been proposed to explain how alcohol ingestion may interfere with retinoid metabolism. First, it has been shown that alcohol may act as a competitive inhibitor of oxidation of retinal to retinoic acid in several organs (e.g. colon and esophagus) [21-23]. Secondly, Liu et al. [24] reported an enhanced catabolism of retinoids into polar metabolites in the liver of ethanol-fed rats compared to pair-fed animals. This was found to be inhibited in vitro and in vivo by chlormethiazol, an inhibitor of cytochrome P4502E1 (CYP2E1). A significant induction of CYP2E1 was found after one week of chronic alcohol ingestion in human subjects [25]. Therefore, the induction of CYP2E1 (and/or other cytochrome P450 isoforms) activity during chronic alcohol intake could add to a reduced bioavailability of retinoic acid subsequently resulting in changes in the expression of RARs and RXRs. Thirdly, it has been shown that levels of cellular retinoic acid binding protein (CRABP) can be reduced in esophageal carcinoma (e.g. adenocarcinoma and SCC) [26]. This, in turn, may lead to alterations of receptor expression and turnover. However, whether these mechanisms play a role in the present study remains to be determined.
RARγ protein levels are elevated in normal esophageal mucosa of patients with SCC
Untill now, information on protein levels of retinoic acid receptors in normal unaffected esophageal tissue has been limited. Recently, in a study determining the role of RXR mRNA in Barett's esophagus mRNA levels of RXRβ were found to be higher in non-malignant tissue in patients with adenoma than in healthy controls [27]. In the present study, protein levels of RARγ were found to be significantly higher in normal esophageal tissue of patients with esophageal SCC in comparison to controls. It may be that the differences between the results of others [27] and the present study are due to differences of patients (e.g. patients with SCC vs. patients with Barrett esophagus) or differences of detection sensitivity (mRNA expression vs. protein).
Conclusion
The results of the present study provide further evidence that a diminished abundance of retinoid acid receptors in esophagus is associated with the development of esophageal SCC in humans. Furthermore, in some patients life style (e.g. smoking and alcohol consumption) may contribute to the alteration of retinoic acid receptor levels in esophagus. Further studies will be required to elucidate this interaction of alcohol consumption, smoking, and retinoic acid receptor signaling with the development of SCC in the esophagus.
Competing interests
The author(s) declare that they have no competing interest.
Authors' contributions
IB has made substantial contributions to the biochemical analysis and the drafting of article. EW has substantially contributed to the acquisition of data, and conception as well as design of the study. EB, AHH, and CB have made substantial contributions to conception and design as well as the interpretation of data. AP has been involved in the design, the drafting of the article, and revised it critically for important intellectual content. All authors have given final approval of the version to be published.
==== Refs
WCRF Food, Nutrition and the Prevention of Cancer: a global perspective American Institute for Cancer Research 1997
Cilley RE Strodel WE Peterson RO Cause of death in carcinoma of the esophagus Am J Gastroenterol 1989 84 147 149 2537007
Bollschweiler E Holscher AH Carcinoma of the esophagus – actual epidemiology in Germany Onkologie 2001 24 180 184 11441301 10.1159/000050312
Bollschweiler E Wolfgarten E Nowroth T Rosendahl U Monig SP Holscher AH Vitamin intake and risk of subtypes of esophageal cancer in Germany J Cancer Res Clin Oncol 2002 128 575 580 12384802 10.1007/s00432-002-0380-z
Valsecchi MG Modelling the relative risk of esophageal cancer in a case-control study J Clin Epidemiol 1992 45 347 355 1569430 10.1016/0895-4356(92)90035-L
Kato I Nomura AM Stemmermann GN Chyou PH Prospective study of the association of alcohol with cancer of the upper aerodigestive tract and other sites Cancer Causes Control 1992 3 145 151 1562704 10.1007/BF00051654
Metzger R Schneider PM Warnecke-Eberz U Brabender J Holscher AH Molecular biology of esophageal cancer Onkologie 2004 27 200 206 15138356 10.1159/000076913
De Luca LM Retinoids and their receptors in differentiation, embryogenesis and neoplasia FASEB 1991 5 2924 2933
Gudas LJ Sporn MB Roberts AB Cellular biology and biochemistry of the retinoids The Retinoids: Biology, Chemistry, and Medicine 1994 Raven Press, New York 443 520
Moon RC Mehta RG Detrisac J Retinoids as chemopreventive agents for breast cancer Cancer Detect Prev 1992 16 73 79 1551141
Xu XC Detection of altered retinoic acid receptor expression in tissue sections using in situ hybridization Histol Histopathol 2001 16 205 212 11193196
Xu M Jin YL Fu J Huang H Chen SZ Qu P Tian HM Li ZY Zhang W The abnormal expression of retinoic acid receptor-beta, p 53 and Ki67 protein in normal, premalignant and malignant esophageal tissues World J Gastroenterol 2002 8 200 202 11925591
Hu L Crowe DL Rheinwald JG Chambon P Gudas LJ Abnormal expression of retinoic acid receptors and keratin 19 by human oral and epidermal squamous cell carcinoma cell lines Cancer Res 1991 51 3972 3981 1713123
Crowe DL Hu L Gudas LJ Rheinwald JG Variable expression of retinoic acid receptor (RAR beta) mRNA in human oral and epidermal keratinocytes; relation to keratin 19 expression and keratinization potential Differentiation 1991 48 199 208 1725165
Lotan R Suppression of squamous cell carcinoma growth and differentiation by retinoids Cancer Res 1994 54 1987s 1990s 8137325
Darwiche N Celli G De Luca LM Specificity of retinoid receptor gene expression in mouse cervical epithelia Endocrinology 1994 134 2018 202 8156902 10.1210/en.134.5.2018
Zhang W Rashid A Wu H Xu XC Differential expression of retinoic acid receptors and p53 protein in normal, premalignant, and malignant esophageal tissues J Cancer Res Clin Oncol 2001 127 237 242 11315258 10.1007/s004320000183
Qiu H Zhang W El Naggar AK Lippman SM Lin P Lotan R Xu XC Loss of retinoic acid receptor-beta expression is an early event during esophageal carcinogenesis Am J Pathol 1999 155 1519 1523 10550308
Phillips DH Smoking-related DNA and protein adducts in human tissues Carcinogenesis 2002 23 1979 2004 12507921 10.1093/carcin/23.12.1979
Song S Xu XC Effect of benzo[a]pyrene diol epoxide on expression of retinoic acid receptor-beta in immortalized esophageal epithelial cells and esophageal cancer cells Biochem Biophys Res Commun 2001 281 872 877 11237740 10.1006/bbrc.2001.4433
Parlesak A Menzl I Bode JC Bode C Inhibtion of retinol oxidation by ethanol in the rat liver and colon Gut 2000 47 825 831 11076882 10.1136/gut.47.6.825
Parlesak A Ellendt K Lindros KO Bode C Acute but not chronic ethanol exposure impairs retinol oxidation in the small and large intestine of the rat Eur J Nutr 2005 44 157 162 15309434 10.1007/s00394-004-0507-x
Crabb DW Pinairs J Hasanadka R Fang M Leo MA Lieber CS Tsukamoto H Motomura K Miyahara T Ohata M Bosron W Sanghani S Kedishvili N Shiraishi H Yokoyama H Miyagi M Ishii H Bergheim I Menzl I Parlesak A Bode C Alcohol and retinoids Alcohol Clin Exp Res 2001 25 207S 217S 11391073 10.1097/00000374-200105051-00034
Liu C Russell RM Seitz HK Wang XD Ethanol enhances retinoic acid metabolism into polar metabolites in rat liver via induction of cytochrome P4502E1 Gastroenterology 2001 120 179 189 11208727 10.1053/gast.2001.20877
Oneta CM Lieber CS Li J Ruttimann S Schmid B Lattmann J Rosman AS Seitz HK Dynamics of cytochrome P4502E1 activity in man: induction by ethanol and disappearance during withdrawal phase J Hepatol 2002 36 47 52 11804663 10.1016/S0168-8278(01)00223-9
Dowlatshahi K Mehta RG Levin B Cerny WL Skinner DB Moon RC Retinoic-acid-binding protein in normal and neoplastic human esophagus Cancer 1984 54 308 311 6327007
Brabender J Lord RV Metzger R Park J Salonga D Danenberg KD Holscher AH Danenberg PV Schneider PM Role of retinoid X receptor mRNA expression in Barrett's esophagus J Gastrointest Surg 2004 8 413 422 15120365 10.1016/j.gassur.2004.02.007
|
16277658
|
PMC1308827
|
CC BY
|
2021-01-04 16:39:19
|
no
|
J Carcinog. 2005 Nov 8; 4:20
|
utf-8
|
J Carcinog
| 2,005 |
10.1186/1477-3163-4-20
|
oa_comm
|
==== Front
J CarcinogJournal of Carcinogenesis1477-3163BioMed Central London 1477-3163-4-221630067310.1186/1477-3163-4-22EditorialNew paradigms, new Hopes: the need for socially responsible research on carcinogenesis Kovvali Gopala [email protected] Gopala Kovvali Ph.D. Editor-in-Chief, Journal of Carcinogenesis, Founder President, Carcinogenesis Foundation, 22 Heritage Drive, Edison, NJ 08820, USA2005 21 11 2005 4 22 22 15 11 2005 21 11 2005 Copyright © 2005 Kovvali; licensee BioMed Central Ltd.2005Kovvali; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
==== Body
It has been three years since the publication of the first article in the Journal of Carcinogenesis; it was the editorial that espoused the need for the launch of the journal. Having seen three springs and three falls, it is time to ask where we are as a journal and where we want to be in the years to come.
The past three years have been satisfying publishing years for JC, given the fact that it is an online publication with processing charges paid by the authors under the open access model and there has been stiff competition from the established print journals. The quality of manuscripts published in JC is very impressive and I am confident that it will only improve as the journal grows. Scientists seem to be conditioned to measure the quality of a journal and the articles it publishes in terms of 'impact factor'. I am glad that JC has made a significant impact on the carcinogenesis research community and I am sure that an impact factor for the journal will follow soon.
As we enter our fourth year, we have a reason to be upbeat. We have been fortunate in attracting several international scientists, both from industry and academia, to the Editorial Board. I am especially pleased that Dr. Rishab Gupta, director of immunodiagnosis and Vice-president for Education at John Wayne Cancer Institute, joined the editorial board as a managing editor. He brings outstanding leadership and vision to the editorial board. I am also pleased to announce that Carcinogenesis Foundation was recently founded with New Jersey, USA, as a home base.
The formation of the Carcinogenesis Foundation is a historic opportunity. While its goal is to promote and advance research in the field of carcinogenesis, it has a specialized mission of understanding the phenomenon of increased cancer incidence among individuals who migrate to the western countries. The phenomenon is conceptualized and can be called Acquired Risk for Cancer Incidence, ARCI. A manuscript being published in the Journal of Carcinogenesis in November 2005 presents an interesting work that supports the concept of ARCI. (Cancer Incidence in the South Asian population of California, 1988–2000 Ratnali V Jain, Paul K Mills and Arti Parikh-Patel). Investigations into new paradigms like ARCI will have a far-reaching impact on the field of carcinogenesis and will open up new horizons for the scientific endeavors that seek to unravel the molecular secrets of malignant transformation of a normal cell. Several interesting questions are currently being addressed in the field of carcinogenesis research and they are expected to lead to insightful answers. Carcinogenesis Foundation and the Journal of Carcinogenesis will have tremendous opportunity to contribute to understanding the carcinogenesis processes and thereby to chemoprevention efforts.
While research is underway in several laboratories around the world, the public is waiting with a hope; a hope that a clear message that they could trust would emerge as to which factors increase the risk of carcinogenesis and which do not as the enlightened scientific community may have already deliberated and researched on them. How difficult is it to answer if vitamin E and B are protective against carcinogenesis or if they promote carcinogenesis? How difficult is it to find out if aspirin or other non-steroidal anti-inflammatory drugs prevent cancers? It may not be difficult if proper scientific methods and approaches are employed and the outcome is not prejudiced one way or the other for any reasons. It may not be difficult if socially conscious scientists approach the questions collectively with a humane outlook. As scientific approach is expected to yield one answer to one question and that answer should be verifiable in any part of the world, where is the source of discordant results to questions like those mentioned above? There seems to be two sources, one is study design and the investigators and the other is inherent in the biology of human subjects and experimental systems used for the study.
The biotechnology industry analysts seem to think that next decade will be dominated by products and processes that address the 'personalized medicine'. It certainly is the ultimate alternative to the current 'one pill cures all' approach. However, it is inevitable to go through community-based medicine before customized medicine can become a reality. Acquired Risk for Cancer Incidence is a concept that entails us to the community based carcinogenesis and cancer prevention research. Carcinogenesis Foundation proposes to participate in and encourage efforts that attempt to understand how the inherent biological and genetic variations in individuals influence the outcome of cancer prevention trials or cancer treatment.
While discordant results in biomedical research may have genesis in genetic diversities of subjects, a distinct contribution for this comes from lack of uniform protocols and procedures and lack of policies to prevent irresponsible scientific practices. Unfortunately, there are no international mechanisms to impose sanctions on ill trained scientists whose results misguide the public as well as the fellow scientists. It is interesting that scientific profession is the only profession that does not have certification and professional regulations like medical profession and others. It is time to think of ways and means by which a unified approach and common protocols are employed in biomedical research. It is also imperative that the individuals who go to the bench are screened and certified for technical competence and scientific integrity. These considerations may call for the creation of a body of a 'jury scientists'. Should cancer researchers be the in the forefront of such an experiment?
|
16300673
|
PMC1308828
|
CC BY
|
2021-01-04 16:39:19
|
no
|
J Carcinog. 2005 Nov 21; 4:22
|
utf-8
|
J Carcinog
| 2,005 |
10.1186/1477-3163-4-22
|
oa_comm
|
==== Front
Cardiovasc DiabetolCardiovascular Diabetology1475-2840BioMed Central London 1475-2840-4-171627447910.1186/1475-2840-4-17Original InvestigationMethylenetetrahydrofolate reductase polymorphism 677C>T is associated with peripheral arterial disease in type 2 diabetes Pollex Rebecca L [email protected] Mary [email protected] Bernard [email protected] Stewart B [email protected] Anthony JG [email protected] Robert A [email protected] Robarts Research Institute, London, Ontario, Canada2 Sandy Lake Health and Diabetes Project, Sandy Lake, Ontario, Canada3 Department of Medicine, University of Toronto and Leadership Sinai Centre for Diabetes, Mount Sinai Hospital, Toronto, Ontario, Canada4 Thames Valley Family Practice Research Unit, University of Western Ontario, London, Ontario, Canada2005 7 11 2005 4 17 17 6 9 2005 7 11 2005 Copyright © 2005 Pollex et al; licensee BioMed Central Ltd.2005Pollex et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Individuals with diabetes are twice as likely to develop peripheral arterial disease (PAD), the manifestation of extensive atherosclerosis throughout the lower extremities. One putative determinant of PAD is the 677C>T polymorphism in the gene encoding methylenetetrahydrofolate reductase (MTHFR), which has previously been found to associate with various diabetic complications including retinopathy, nephropathy, atherosclerosis and coronary heart disease. The objective of this study was to investigate a possible role for the MTHFR 677C>T gene polymorphism with PAD in subjects with type 2 diabetes from an isolated aboriginal Canadian population.
Methods
The 677C>T MTHFR gene polymorphism was genotyped in 138 subjects of Oji-Cree descent. Participants were selected from a community-wide survey that included PAD assessment by ankle-brachial index (ABI) measurement, and also intermittent claudication assessment by the Rose questionnaire.
Results
MTHFR 677T allele carriers had an increased risk of PAD with an odds ratio of 3.54 (95% CI 1.01, 12.4), P = 0.049, after adjustment for age, sex, duration of diabetes, hypertension, current smoking habits, and use of insulin or oral treatment for diabetes. None of these additional co-variables was significantly associated with PAD. No association was found between MTHFR genotype and intermittent claudication.
Conclusion
The genetic influence of the MTHFR 677C>T genotype on diabetic PAD is modest, yet for the Oji-Cree it is a major risk factor in comparison to other traditional risk factors.
==== Body
Background
Peripheral arterial disease (PAD) of the lower extremities, due to atherosclerosis occurring at the aortic bifurcation, femoral and popliteal arteries, commonly manifests as asymptomatic changes in intermediate phenotypes such as the ankle-brachial index (ABI). The most common symptom experienced is the aches and pains of intermittent claudication, and in extreme situations, individuals may develop gangrene and require lower-limb amputations [1]. While most subjects with PAD remain asymptomatic, all have markedly increased risk of developing cardiovascular disease, as PAD is fundamentally a sign of systemic atherosclerosis. Compared with age-matched controls, patients with intermittent claudication have a threefold increase in cardiovascular mortality [2,3] and a low ABI has been shown to be an independent predictor of both all-cause and cardiovascular mortality [4,5]. The presence of diabetes is associated with a doubling of PAD risk [6]; up to 15% of subjects with type 2 diabetes can have clinically significant PAD [6,7]. Other well-known risk factors for PAD include hypertension, dyslipidemia, and smoking [1].
Efforts to determine some of the genetic factors underlying susceptibility to PAD have met limited success [8-14]. Candidate genes include those that have shown association with atherosclerosis in other vascular beds. One such candidate is the gene encoding methylenetetrahydrofolate reductase (MTHFR), a key enzyme in the alternative pathway of homocysteine metabolism that remethylates homocysteine to methionine. Elevated plasma homocysteine has been reported among carriers of the MTHFR 677C>T (Ala222Val; MIM 607093.0003) thermolabile single nucleotide polymorphism (SNP) [15]. This dysfunctional SNP is associated with reduced enzyme activity, resulting in a relative deficiency in the remethylation process [15], leading to elevated plasma homocysteine. Elevated plasma homocysteine concentration appears to be significantly associated with PAD [9,16]. Hyperhomocysteinemia may promote vascular disease through endothelial injury, predisposing the vessel to atherosclerosis [17]. Since the MTHFR 677C>T SNP is an important determinant of plasma homocysteine concentration, this polymorphism may represent an important genetic risk factor in vascular disease.
The MTHFR 677C>T SNP has previously been found to associate with various diabetic complications, including retinopathy [18-21], nephropathy [22-25], atherosclerosis [26], and coronary heart disease [27]. However, no studies to date have reported an association specifically for PAD, as measured, for instance, by the non-invasive ABI, a tool that has proven its usefulness in predicting future cardiovascular events [28]. Thus the objective of this study was to investigate a possible role for the MTHFR 677C>T polymorphism with PAD (ABI) in subjects with type 2 diabetes.
Research design and methods
Study Sample
Patients in this study were participants in the Sandy Lake Complications Prevalence and Risk Factor Study, which was initiated to study the prevalence of, and risk factors for, complications of type 2 diabetes in aboriginal Canadians [29]. Sandy Lake, Ontario, is a remote Oji-Cree community, found at the 55th parallel of latitude, in the subarctic boreal forest of central Canada. For this community-based, cross-sectional study, 189 eligible subjects with type 2 diabetes were enrolled, although the sample size varied for some variables given time-limited access to certain diagnostic equipment. Of these, 173 individuals had available DNA for MTHFR 677C>T genotyping. For the determination of PAD, 140 subjects participated, with an overlap of 138 subjects who had both ABI and MTHFR 677C>T genotype determination. Signed informed consent was obtained from all participants. The study was approved by both the Sandy Lake First Nation Band Council and the Mount Sinai Hospital Ethics Review Committee.
Clinical characteristics and biochemical analysis
Measurements of fasting blood analytes, including glucose, lipid and lipoprotein concentrations, and percent glycoslyated hemoglobin (HbA1c), were performed as described [30]. Standardized procedures were used to measure blood pressure, height, weight, and waist circumferences [30]. Information on diabetes duration, diabetes treatment, and tobacco use, was obtained from interviewer-administered questionnaires [30]. Hypertensive individuals were defined as those subjects having either blood pressure exceeding systolic 130 mmHg and/or diastolic 80 mmHg, and/or receiving antihypertensive treatment.
Diagnosis of peripheral arterial disease and intermittent claudication
The diagnosis of PAD was based on the ABI, which was measured using a blood pressure cuff and Doppler stethoscope. Systolic blood pressure was assessed at 3 sites on each side (brachial, posterior tibial, and dorsalis pedis arteries). Left and right ABI measurements were obtained by selecting the highest leg systolic blood pressure reading (either the posterior tibial systolic blood pressure or the dorsalis pedis systolic blood pressure) and dividing it by the mean brachial systolic blood pressure. The same individual performed all measurements. The use of ABI as a screening test for atherosclerotic disease has been previously validated [31], and test results have been shown to have high reproducibility when performed by trained professionals [32]. By convention, PAD was defined as an ABI of less than 0.95; subjects with an ABI greater than 1.40 were considered to have non-compressible vessels [29].
Intermittent claudication was assessed using the Rose (World Health Organization) questionnaire, with a positive score indicating the presence of leg pain, using the provided standard algorithm for diagnosis [33]. Leg pain was not considered intermittent claudication if it started when standing still or sitting, if it did not include the calves, or if it did not occur when walking up hill or hurrying [33].
Genotyping MTHFR 677C>T
An established procedure was used to genotype the MTHFR 677C>T SNP [15]. Briefly, exon 4 was amplified using the following primers: 5'-CAA AGG CCA CCC CGA AGC and 5'-AGG ACG GTG CGG TGA GAG TG. Samples were amplified for 30 cycles, each of which consisted of denaturing at 94°C for 30 s, annealing at 58°C for 30 s, and extension at 72°C for 30 s. After HinfI (New England Biolabs, Mississauga, Ontario, Canada) digestion of the resulting 245 bp fragment, the C allele yielded only a single 245 bp fragment, and the T allele yielded two fragments with sizes 176 and 69 bp. Electrophoresis in a 2.5% agarose gel followed by ethidium bromide staining and ultraviolet illumination allowed detection of the alleles.
Statistical analysis
SAS version 8.2 (SAS Institute, Cary, NC) was used for all statistical comparisons. Data are presented as means ± standard deviation (s.d.) or as percentages for categorical variables. Logarithmic transformations (natural log) were used if data were not normally distributed. The transformed variables were used for parametric statistical analyses, but the untransformed values are presented in the tables. For continuous variables, differences between the groups were tested by the Student's t test; categorical variables were tested by χ2 analysis or by Fisher's exact test. PAD was analyzed by multivariate logistic regression, with MTHFR 677C>T genotype, duration of diabetes, current smoking habits, blood pressure, antihypertensive treatment, treatment for diabetes (insulin or oral agent), age, and sex included as independent variables. Deviation of genotype frequencies from those predicted by Hardy-Weinberg law was tested by χ2 analysis. All statistical tests were two-sided and statistical significance was taken at nominal P < 0.05 for all comparisons.
Results
Characteristics of Oji-Cree type 2 diabetic patients
Study participants (35.5% male) had an average age of 46.7 ± 11.2 years for males and 45.8 ± 11.9 years for females, and a mean duration of diabetes of 8.27 ± 5.49 years for males and 8.90 ± 6.98 years for females. In comparison to males, females had significantly higher BMI (31.1 ± 5.2 vs 29.0 ± 4.5 kg/m2, P = 0.020), and a lower frequency of hypertension (55.1% vs 77.6%, P = 0.0088). Approximately 14% had PAD, with PAD more frequent on the right side than the left (~12% vs ~8%), and no significant difference between the numbers of males and females affected (14.3% males vs 14.6% females, P=NS [0.96]). Slightly more than 1 out of 20 subjects had intermittent claudication (6.12% males vs 62% females, P = NS [1.00]). Clinical attributes of the subjects, according to the presence or absence of PAD, are shown in Table 1. Significant differences between those with and without PAD were noted for total cholesterol concentrations and systolic blood pressure, with PAD-affected individuals having, on average, lower total cholesterol (P = 0.036) and higher systolic blood pressure (P = 0.0028). There were also fewer affected individuals taking antidiabetic drugs (P = 0.045).
Table 1 Characteristics of Oji-Cree type 2 diabetic patients according to the presence or absence of PAD (max n = 138)
PAD present PAD absent P
N N
Sex (% male) 20 35.0 118 35.6 NS (0.96)
Age (years) 20 48.0 ± 11.2 118 45.8 ± 11.7 NS (0.44)
BMI (kg/m2) 20 30.4 ± 5.7 117 30.3 ± 4.9 NS (0.98)
Waist circumference (cm) 20 102 ± 13 117 104 ± 9 NS (0.52)
Current smokers (%) 20 65.0 118 47.5 NS (0.15)
Duration of diabetes (years) 20 8.70 ± 4.64 118 8.67 ± 6.76 NS (0.98)
Insulin treatment (%) 16 25.0 98 18.4 NS (0.51)
Intake of antidiabetic drugs (%) 16 31.3 98 58.2 0.045
Taking lipid medication (%) 20 5.0 118 16.1 NS (0.31)
Fasting glucose (mmol/L) 19 10.27 ± 2.79 111 9.96 ± 3.90 NS (0.48)
HbA1c (%) 20 8.30 ± 1.93 116 8.44 ± 2.29 NS (0.92)
Total cholesterol (mmol/L) 19 4.43 ± 0.63 114 4.93 ± 1.06 0.036
Plasma triglycerides (mmol/L) 19 1.54 ± 0.60 114 2.18 ± 2.68 NS (0.059)
Systolic BP (mmHg) 20 136 ± 21 118 123 ± 16 0.0028
Diastolic BP (mmHg) 20 74 ± 10 118 75 ± 9 NS (0.81)
Hypertensive (%) 20 70.0 118 61.9 NS (0.49)
CRP (mg/L) 19 6.86 ± 12.0 114 4.97 ± 5.87 NS (0.99)
ABI, left 20 0.96 ± 0.12 118 1.09 ± 0.08 <0.0001
ABI, right 20 0.93 ± 0.06 118 1.10 ± 0.08 <0.0001
Intermittent claudication (%) 20 15.0 118 4.2 NS (0.091)
Data are means ± s.d., unless otherwise indicated.
PAD, peripheral arterial disease: ABI <0.95; Hypertension: systolic BP ≥ 130 and/or diastolic BP ≥ 80 and/or antihypertensive treatment.
Abbreviations: BMI, body mass index; HbA1c, glycosylated hemoglobin; BP, blood pressure; CRP, C-reactive protein; ABI, ankle-brachial index; NS, not significant.
The overall allele frequencies for the 677C and 677T alleles were 0.89 and 0.11, respectively. All 677T carriers were heterozygotes (677C/T); no 677T/T homozygotes were found in the population sample studied. There was no significant difference found for the 677C>T genotype frequency between sex, with 16.3% of males and 24.7% of females, respectively, being carriers of the 677T allele (P = NS [0.25]). There was no significant deviation of the genotype frequencies, overall or according to sex, from those predicted by Hardy-Weinberg law.
Characteristics of subjects according to the presence or absence of the 677T allele are presented in Table 2. The thirty 677T carriers had significantly elevated BMI (32.6 ± 5.5 vs 29.7 ± 4.7 kg/m2, P = 0.0050), and higher systolic blood pressure (132 ± 22 vs 123 ± 16 mmHg, P = 0.021), in comparison to 677C/C homozygotes. There was no difference between 677T carriers and noncarriers regarding age, waist circumference, % current smoker, duration of diabetes, % hypertensive, diastolic blood pressure, mean ABI, and plasma concentrations of fasting blood glucose, HbA1c, total cholesterol, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, triglycerides, and C-reactive protein.
Table 2 Characteristics of Oji-Cree type 2 diabetic patients according to MTHFR 677C>T genotype (max n = 138)
N 677C/C 677C/T P
N (men/women) 138 108 (41/67) 30 (8/22) NS (0.25)
Age (years) 138 46.4 ± 11.1 45.1 ± 13.5 NS (0.60)
BMI (kg/m2) 137 29.7 ± 4.7 32.6 ± 5.5 0.0050
Waist circumference (cm) 137 103 ± 9 106 ± 11 NS (0.063)
Current smoker (%) 138 50.0 50.0 NS (1.00)
Duration of diabetes (years) 138 8.53 ± 6.35 9.20 ± 7.01 NS (0.62)
Fasting glucose (mmol/L) 130 10.20 ± 3.91 9.28 ± 3.03 NS (0.33)
HbA1c (%) 136 8.36 ± 2.25 8.66 ± 2.18 NS (0.47)
Total cholesterol (mmol/L) 133 4.91 ± 1.07 4.67 ± 0.82 NS (0.27)
HDL-C (mmol/L) 133 1.21 ± 0.29 1.21 ± 0.29 NS (0.91)
LDL-C (mmol/L) 130 2.75 ± 0.71 2.70 ± 0.67 NS (0.78)
Plasma triglycerides (mmol/L) 133 2.20 ± 2.78 1.66 ± 0.60 NS (0.15)
Systolic BP (mmHg) 138 123 ± 16 132 ± 22 0.021
Diastolic BP (mmHg) 138 73.9 ± 8.9 77.4 ± 9.4 NS (0.077)
Hypertensive (%) 138 62.0 66.7 NS (0.64)
CRP (mg/L) 133 4.91 ± 6.65 6.53 ± 8.37 NS (0.19)
ABI, left 138 1.08 ± 0.09 1.05 ± 0.12 NS (0.26)
ABI, right 138 1.08 ± 0.09 1.07 ± 0.13 NS (0.73)
Data are means ± s.d., unless otherwise indicated.
Hypertension: systolic BP ≥ 130 and/or diastolic BP ≥ 80 and/or antihypertensive treatment
Abbreviations: BMI, body mass index; HbA1c, glycosylated hemoglobin; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; BP, blood pressure; CRP, C-reactive protein; ABI, ankle-brachial index; NS, not significant.
Association of 677C>T SNP with peripheral arterial disease
Significantly more individuals carrying the 677T allele had PAD, as diagnosed by ABI (26.7% vs 11.1%, P = 0.042); particularly when measured on the right side (23.3% vs 8.33%, P = 0.047) (Table 3). However, no significant difference between carriers and non-carriers was noted for symptomatic intermittent claudication as determined by the Rose questionnaire [6.67% vs 5.56%, P = NS (0.68)].
Table 3 PAD and intermittent claudication according to MTHFR 677C>T genotype
677C/C n = 108 677C/T n = 30 OR (95% CI) P*
PAD, left or right (n, %) 12 (11.1%) 8 (26.7%) 2.91 (1.06, 7.97) 0.042
PAD, left (n, %) 7 (6.48%) 4 (13.3%) 2.22 (0.60, 8.16) NS (0.25)
PAD, right (n, %) 9 (8.33%) 7 (23.3%) 3.35 (1.13, 9.93) 0.047
Intermittent claudication (n, %) 6 (5.56%) 2 (6.67%) 1.20 (0.26, 5.64) NS (0.68)
* Fisher's exact test
Abbreviations: OR, odds ratio; PAD, peripheral arterial disease; NS, not significant.
Multivariate logistic regression (Table 4) revealed that MTHFR 677C>T genotype was still significantly associated, albeit marginally, with increased risk of PAD [OR 3.54 (1.01, 12.4), P = 0.049], after adjustment for age, sex, duration of diabetes, hypertension, current smoking habits, and use of insulin or oral treatment for diabetes. None of these additional co-variables was significantly associated with PAD.
Table 4 Multivariate associations of risk factors with PAD
Dependent variable Independent variable Unit OR (95% CI) P
PAD, left or right MTHFR 677C>T 677C/T 3.54 (1.01, 12.4) 0.049
oral treatment for diabetes + 0.31 (0.082, 1.20) NS (0.091)
current smoker + 3.04 (0.82, 11.2) NS (0.095)
hypertension + 2.58 (0.58, 11.6) NS (0.22)
age 10 yr 1.36 (0.80, 2.29) NS (0.26)
duration of diabetes 5 yr 0.79 (0.45, 1.38) NS (0.41)
insulin administration + 0.96 (0.18, 5.08) NS (0.97)
sex male 1.00 (0.30, 3.34) NS (0.99)
Hypertension: systolic blood pressure ≥ 130 and/or diastolic blood pressure ≥ 80 and/or antihypertensive treatment.
Abbreviations: OR, odds ratio; PAD, peripheral arterial disease; MTHFR, methylenetetrahydrofolate reductase; NS, not significant.
Discussion
In Canadian Oji-Cree subjects with type 2 diabetes, we found that: 1) MTHFR 677T carriers had an increased risk of PAD [OR 3.54 (1.01, 12.4), P = 0.049], adjusting for age, sex, duration of diabetes, hypertension, current smoking habits, and diabetes treatment; and 2) there was no significant association between MTHFR genotype and intermittent claudication, a much more advanced stage of PAD.
Our finding of a significant association between MTHFR genotype and PAD for subjects with type 2 diabetes is novel, to our knowledge. Only one other study to date has examined the relationship between PAD and MTHFR genotype in subjects with type 2 diabetes. In 135 patients with PAD and 219 controls, Ciccarone et al. found that though elevated homocysteine levels associated with the severity of PAD, as assessed by colour-duplex ultrasound, there was no significant association between MTHFR genotype and PAD [9]. Our observed association is of borderline statistical significance, yet given that under multivariate analysis MTHFR genotype was the most significant risk factor in comparison to other traditional risk factors, including smoking, hypertension, and duration of diabetes, investigation into its potential importance is warranted.
Though not significant, a 3-fold increased risk of PAD was observed for subjects who were current smokers. This finding stresses the importance of smoking cessation programmes in aboriginal communities, especially given the high prevalence of smoking in this population (~50% current smokers for both males and females). Taking preventative measures now is also critical considering the relatively young age of the population. Though the prevalence of PAD was ~14%, and the prevalence of intermittent claudication was only ~6%, the mean age of the study participants was under 50, younger than the majority of studies involving PAD [9,10,13]. A full expression of disease phenotypes associated with diabetes for the Oji-Cree is yet to be seen.
A limitation of our study was the small sample size. With less than 200 subjects included in each analysis, the power to detect significant associations was low. Most notably, the observation of no significant association for any of the well-known risk factors in the logistic regression analysis for PAD was likely due to the small sample size. In spite of the small sample size, the observation of an OR of greater than 3.5 for the MTHFR genotype and PAD suggests an association of notably strong magnitude which would be of interest to confirm in future studies.
Our results are also limited by the absence of both dietary information and plasma folate and homocysteine concentrations for our study subjects. Previous studies have found the prevalence of inadequate dietary folate intake in the Oji-Cree to be 37%, which is more than twice the average for the general Canadian population, and may contribute to compromised MTHFR enzyme activity [34]. These studies, however, took place in an era prior to the supplementation of the food source with folate, and thus may be an overestimation of the present situation. Further detailed clinical measurements would have helped to provide insight on the relationships between MTHFR genotype, plasma homocysteine concentration, and PAD. Other studies, however, have found significant associations for homocysteine levels with increased atherosclerosis, yet no significant association for MTHFR genotype [9,35], raising the question of possible alternative causes for hyperhomocysteinemia and questioning the real importance of the MTHFR genotype. Nonetheless, our data are consistent with the concept that MTHFR genotype may have a mechanistic role in the development and progression of atherosclerosis in the lower extremities.
Conclusion
In conclusion, we have observed that the presence of the MTHFR 677T allele was significantly associated with an ~3.5-fold increased risk for PAD, as assessed by ABI, in subjects with type 2 diabetes. The significance of the relationship between the MTHFR SNP and ABI was greater than that of any other common risk factor, including age, sex, duration of diabetes, hypertension, current smoking habits, or diabetes treatment. Since this variant is common, it could represent an important genetic determinant of PAD risk in the general population.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
RLP participated in the design of the study, genotyping, analysis of the data, and writing of the manuscript. MM participated in the on-site data collection and preformed all blood pressure measurements. AJGH, BZ, and SBH provided patients and data for the study, and assisted with manuscript revisions. RAH participated in the design of the study and writing of the manuscript. All authors read and approved the final manuscript.
Acknowledgements
The authors gratefully acknowledge the chief, council and community members of Sandy Lake First Nation whose partnership and co-operation was essential in the design and implementation of this project. RLP is supported by a Natural Sciences and Engineering Research Council of Canada Graduate Scholarship and the Canadian Institutes of Health Research Strategic Training Program in Vascular Research. BZ holds the Sam and Judy Pencer Family Chair in Diabetes Research. SBH holds the Ian McWhinney Chair for Studies in Family Medicine at the Schulich School of Medicine, University of Western Ontario. AJGH is supported through a Canadian Diabetes Association Research Scholarship and a University of Toronto Banting and Best Diabetes Centre New Investigator Award. RAH holds a Canada Research Chair (Tier I) and is a Career Investigator of the Heart and Stroke Foundation of Ontario. The project was supported by operating grants from the Canadian Diabetes Association and the Ontario Research and Development Challenge Fund (99-0507).
==== Refs
Peripheral arterial disease in people with diabetes Diabetes Care 2003 26 3333 3341 14633825
Jelnes R Gaardsting O Hougaard Jensen K Baekgaard N Tonnesen KH Schroeder T Fate in intermittent claudication: outcome and risk factors Br Med J (Clin Res Ed) 1986 293 1137 1140 3094806
Reunanen A Takkunen H Aromaa A Prevalence of intermittent claudication and its effect on mortality Acta Med Scand 1982 211 249 256 7102362
Newman AB Sutton-Tyrrell K Vogt MT Kuller LH Morbidity and mortality in hypertensive adults with a low ankle/arm blood pressure index Jama 1993 270 487 489 8147959 10.1001/jama.270.4.487
Vogt MT Cauley JA Newman AB Kuller LH Hulley SB Decreased ankle/arm blood pressure index and mortality in elderly women Jama 1993 270 465 469 8320785 10.1001/jama.270.4.465
Selvin E Erlinger TP Prevalence of and risk factors for peripheral arterial disease in the United States: results from the National Health and Nutrition Examination Survey, 1999-2000 Circulation 2004 110 738 743 15262830 10.1161/01.CIR.0000137913.26087.F0
Wattanakit K Folsom AR Selvin E Weatherley BD Pankow JS Brancati FL Hirsch AT Risk factors for peripheral arterial disease incidence in persons with diabetes: the Atherosclerosis Risk in Communities (ARIC) Study Atherosclerosis 2005 180 389 397 15910867 10.1016/j.atherosclerosis.2004.11.024
Carter AM Mansfield MW Grant PJ Polymorphisms of platelet glycoproteins in relation to macrovascular disease in type 2 diabetes mellitus Diabet Med 1998 15 315 319 9585397 10.1002/(SICI)1096-9136(199804)15:4<315::AID-DIA590>3.3.CO;2-1
Ciccarone E Di Castelnuovo A Assanelli D Archetti S Ruggeri G Salcuni N Donati MB Capani F Iacoviello L Homocysteine levels are associated with the severity of peripheral arterial disease in Type 2 diabetic patients J Thromb Haemost 2003 1 2540 2547 14675090 10.1111/j.1538-7836.2003.00500.x
Hamanishi T Furuta H Kato H Doi A Tamai M Shimomura H Sakagashira S Nishi M Sasaki H Sanke T Nanjo K Functional variants in the glutathione peroxidase-1 (GPx-1) gene are associated with increased intima-media thickness of carotid arteries and risk of macrovascular diseases in japanese type 2 diabetic patients Diabetes 2004 53 2455 2460 15331559
Koppel H Krippl P Gasser R Wascher TC Paulweber B Pilger E Renner W Hemochromatosis gene (HFE) polymorphisms are not associated with peripheral arterial disease Thromb Haemost 2004 91 1258 1259 15175819
Pettersson-Fernholm K Forsblom C Hudson BI Perola M Grant PJ Groop PH The functional -374 T/A RAGE gene polymorphism is associated with proteinuria and cardiovascular disease in type 1 diabetic patients Diabetes 2003 52 891 894 12606536
Thomas GN Critchley JA Tomlinson B Cockram CS Chan JC Peripheral vascular disease in Type 2 diabetic Chinese patients: associations with metabolic indices, concomitant vascular disease and genetic factors Diabet Med 2003 20 988 995 14632699 10.1046/j.1464-5491.2003.01046.x
Tseng CH Tseng CP Lack of association between angiotensin-converting enzyme gene polymorphism and peripheral vascular disease in type 2 diabetic patients in Taiwan Circ J 2002 66 1014 1018 12419932 10.1253/circj.66.1014
Frosst P Blom HJ Milos R Goyette P Sheppard CA Matthews RG Boers GJ den Heijer M Kluijtmans LA van den Heuvel LP A candidate genetic risk factor for vascular disease: a common mutation in methylenetetrahydrofolate reductase Nat Genet 1995 10 111 113 7647779 10.1038/ng0595-111
Malinow MR Plasma homocyst(e)ine and arterial occlusive diseases: a mini-review Clin Chem 1995 41 173 176 7813076
Refsum H Ueland PM Nygard O Vollset SE Homocysteine and cardiovascular disease Annu Rev Med 1998 49 31 62 9509248 10.1146/annurev.med.49.1.31
Maeda M Yamamoto I Fukuda M Nishida M Fujitsu J Nonen S Igarashi T Motomura T Inaba M Fujio Y Azuma J MTHFR gene polymorphism as a risk factor for diabetic retinopathy in type 2 diabetic patients without serum creatinine elevation Diabetes Care 2003 26 547 548 12547903
Neugebauer S Baba T Kurokawa K Watanabe T Defective homocysteine metabolism as a risk factor for diabetic retinopathy Lancet 1997 349 473 474 9040583 10.1016/S0140-6736(05)61185-3
Sun J Xu Y Zhu Y Lu H Deng H Fan Y Sun S Zhang Y The relationship between MTHFR gene polymorphisms, plasma homocysteine levels and diabetic retinopathy in type 2 diabetes mellitus Chin Med J (Engl) 2003 116 145 147 12667409
Zintzaras E Chatzoulis DZ Karabatsas CH Stefanidis I The relationship between C677T methylenetetrahydrofolate reductase gene polymorphism and retinopathy in type 2 diabetes: a meta-analysis J Hum Genet 2005 50 267 275 15902512 10.1007/s10038-005-0250-z
Ksiazek P Bednarek-Skublewska A Buraczynska M The C677T methylenetetrahydrofolate reductase gene mutation and nephropathy in type 2 diabetes mellitus Med Sci Monit 2004 10 BR47 51 14737040
Moczulski D Fojcik H Zukowska-Szczechowska E Szydlowska I Grzeszczak W Effects of the C677T and A1298C polymorphisms of the MTHFR gene on the genetic predisposition for diabetic nephropathy Nephrol Dial Transplant 2003 18 1535 1540 12897091 10.1093/ndt/gfg211
Sun J Xu Y Zhu Y [The association of methylenetetrahydrofolate reductase gene polymorphism with nephropathy in type 2 diabetes mellitus in Chinese] Zhonghua Nei Ke Za Zhi 2001 40 529 532 11718053
Sun J Xu Y Zhu Y Lu H Genetic polymorphism of methylenetetrahydrofolate reductase as a risk factor for diabetic nephropathy in Chinese type 2 diabetic patients Diabetes Res Clin Pract 2004 64 185 190 15126006 10.1016/j.diabres.2003.10.022
Hasegawa G Obayashi H Kamiuchi K Nakai M Kanatsuna T Yamaguchi M Tanaka T Shigeta H Fujii M Yoshikawa T Nakamura N The association between end-stage diabetic nephropathy and methylenetetrahydrofolate reductase genotype with macroangiopathy in type 2 diabetes mellitus Exp Clin Endocrinol Diabetes 2003 111 132 138 12784186 10.1055/s-2003-39785
Sun J Xu Y Xue J Zhu Y Lu H Methylenetetrahydrofolate reductase polymorphism associated with susceptibility to coronary heart disease in Chinese type 2 diabetic patients Mol Cell Endocrinol 2005 229 95 101 15607533 10.1016/j.mce.2004.09.003
Doobay AV Anand SS The Sensitivity and Specificity of the Ankle-Brachial Index to Predict Future Cardiovascular Outcomes. A Systematic Review Arterioscler Thromb Vasc Biol 2005 15879302
Hanley AJG Harris SB Mamakeesick M Goodwin K Fiddler E Hegele RA McLaughlin JR Zinman B Complications of type 2 diabetes among Native Canadians: increasing our understanding of prevalence and risk factors Canadian J Diabetes 2003 27 455 463
Hanley AJ Harris SB Mamakeesick M Goodwin K Fiddler E Hegele RA Spence JD House AA Brown E Schoales B McLaughlin JR Klein R Zinman B Complications of Type 2 Diabetes Among Aboriginal Canadians: Prevalence and associated risk factors Diabetes Care 2005 28 2054 2057 16043760
Shinozaki T Hasegawa T Yano E Ankle-arm index as an indicator of atherosclerosis: its application as a screening method J Clin Epidemiol 1998 51 1263 1269 10086818 10.1016/S0895-4356(98)00122-X
Matzke S Franckena M Alback A Railo M Lepantalo M Ankle brachial index measurements in critical leg ischaemia--the influence of experience on reproducibility Scand J Surg 2003 92 144 147 12841555
Rose GA Blackburn H Cardiovascular survey methods Monogr Ser World Health Organ 1968 56 1 188 4972212
Wolever TMS Hamad S Gittelsohn J Hanley AJG Logan A Harris SB Zinman B Nutrient intake and food use in an Ojibwa-Cree community in northern Ontario assessed by 24 h dietary recall Nutr Res 1997 17 603 618 10.1016/S0271-5317(97)00033-X
Spence JD Malinow MR Barnett PA Marian AJ Freeman D Hegele RA Plasma homocyst(e)ine concentration, but not MTHFR genotype, is associated with variation in carotid plaque area Stroke 1999 30 969 973 10229729
|
16274479
|
PMC1308829
|
CC BY
|
2021-01-04 16:38:10
|
no
|
Cardiovasc Diabetol. 2005 Nov 7; 4:17
|
utf-8
|
Cardiovasc Diabetol
| 2,005 |
10.1186/1475-2840-4-17
|
oa_comm
|
==== Front
Cell Commun SignalCell communication and signaling : CCS1478-811XBioMed Central London 1478-811X-3-141630305110.1186/1478-811X-3-14ResearchActivation of nuclear factor kappa B (NF-κB) by connective tissue growth factor (CCN2) is involved in sustaining the survival of primary rat hepatic stellate cells Gao Runping [email protected] David R [email protected] Center for Cell and Vascular Biology, Children's Research Institute, Columbus Ohio 43205 USA2 Department of Surgery, The Ohio State University, Columbus, Ohio 43212 USA3 Department of Molecular and Cellular Biochemistry, The Ohio State University, Columbus, Ohio 43212 USA2005 22 11 2005 3 14 14 29 8 2005 22 11 2005 Copyright © 2005 Gao and Brigstock; licensee BioMed Central Ltd.2005Gao and Brigstock; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background/Aims
Connective tissue growth factor (CCN2) is a matricellular protein that plays a role in hepatic stellate cell (HSC)-mediated fibrogenesis. The aim of this study was to investigate the regulation by CCN2 of cell survival pathways in primary HSC.
Methods
Primary HSC were obtained by in situ enzymatic perfusion of rat liver. NF-κB activation was assessed by immunoblotting for IκBα phosphorylation and degradation and by NF-κB p50 or p65 nuclear accumulation. NF-κB DNA-binding activity was determined by gel mobility shift assay while NF-κB response gene expression was evaluated using a luciferase reporter. Cell viability was assessed by Trypan blue staining or ATP luminescent assay while apoptosis was evaluated by caspase-3 activity.
Results
CCN2 induced IκBα phosphorylation and degradation as well as nuclear accumulation of NF-κB. Activated NF-κB comprised three dimers, p65/p65, p65/p50 and p50/p50, that individually bound to DNA-binding sites and subsequently triggered transcriptional activity. This was confirmed by showing that CCN2 promoted activity of a NF-κB luciferase reporter. CCN2 promoted survival of serum-starved HSC and protected the cells from death induced by blocking the NF-κB signaling pathway using Bay-11-7082, a specific inhibitor of IκBα phosphorylation.
Conclusion
CCN2 contributes to the survival of primary HSC through the NF-κB pathway.
connective tissue growth factorCCN2hepatic stellate cellNF-κBsurvivalapoptosisfibrosis
==== Body
Introduction
Hepatic stellate cells (HSC) are the primary targets of fibrogenic stimuli in the injured liver. During the development of fibrosis, HSC undergo a transition from resting vitamin A-rich cells to an activated myofibroblastic phenotype characterized by loss of vitamin A, expression of α-smooth muscle actin, enhanced proliferation and increased production of various extracellular matrix components [1-4]. Activation of HSC has been identified as a central event in hepatic fibrosis and is regulated by a wide variety of molecules including cytokines, cell-surface receptors, signal transduction molecules and factors that regulate HSC gene expression at the transcriptional and post-transcriptional levels [3-6].
Connective tissue growth factor (CCN2, also known as CTGF) is a cysteine-rich matricellular protein that regulates cell adhesion, migration, proliferation, survival, and differentiation [7]. It has fibrogenic properties in vitro and is over-expressed in many fibrotic lesions, including those of the skin, lung, kidney and liver [8-12]. CCN2 production is enhanced during progressive activation of primary rat HSC in vitro as well as by transforming growth factor-β [11-13]. CCN2 induces migration, proliferation and adhesion of HSC as well as enhanced expression of type I collagen [14-19].
Transcription factor NF-κB is a key regulator of the growth, differentiation, and fate of mammalian cells [20]. NF-κB exists in virtually all cell types and represents a family of inducible transcription factors that are activated by a variety of stimuli including viral infection, lipopolysaccharide, oxidative stress, and cytokines [21]. The active form of NF-κB is found in the nucleus as either a heterodimer or a homodimer composed of five members of the Rel family of proteins (p65, p50, p52, c-Rel, and RelB) [20]. It has recently been reported that transcriptional repressor CBF1 plays a key role in regulating NF-κB activity through its interaction with a dual NF-κB/CBF1-binding site in the IκBα promoter [22,23]. IκBα regulates NF-κB activity by directly interacting with the transcriptional factor to form inactive complexes that are located to the cytoplasm. Following specific signaling, phosphorylation of IκBα at serine 32 and 36 by IκB kinase leads to its ubiquitinylation and degradation by the proteasome, and transport of active NF-κB to nucleus [24]. Active NF-κB is involved in the expression of numerous cytokines, acute phase response proteins, adhesion molecules and Rel/IκB proteins [20]. Also, when NF-κB activation is prevented or inhibited, cells undergo enhanced apoptosis showing that active NF-κB exerts a cytoprotective role by inhibiting apoptosis [25].
Several studies have compared NF-κB activity in quiescent versus activated HSC [26-28]. NF-κB activity is increased in cultured activated HSC but it is not required for either cell proliferation or the process of activation. In contrast, active NF-κB plays an important role in preventing apoptosis of activated HSC [26,29]. Understanding mechanisms of HSC survival may provide the basis for novel anti-fibrotic therapies that focus on the ability to clear activated HSC from the liver by inducing them to undergo apoptosis. Since the principal CCN2 receptor on HSC is integrin αvβ3 [19], which is intimately associated with HSC survival [30], we have investigated the role of CCN2 in NF-κB activation and HSC survival.
Results
CCN2 induces phosphorylation of IκBα and translocation of NF-κB
In most cell types, NF-κB is found in the cytoplasm as an inactive dimer bound to one of the IκB inhibitory proteins (IκBα, IκBβ, or IκBγ) that mask its nuclear localization signal. As assessed by Western blotting of cytoplasmic protein extracts day 4 primary HSC, phospho-IκBα was elevated while total IκBα was decreased following stimulation by CCN2 (Figure 1A), indicating that CCN2 could induce IκBα phosphorylation and degradation. Additionally, following CCN2 stimulation, levels of p65 and p50 were reduced in the cytoplasm but increased in the nucleus (Figure 1B), consistent with the notion that CCN2-induced IκBα phosphorylation and degradation was associated with translocation of cytoplasmic NF-κB to the nucleus.
Figure 1 Effect of CCN2 on stimulation of IκBα phophorylation and NF-κB translocation. Freshly isolated rat HSC were cultured for 24 h in 5% FBS DMEM and for another 48 h in serum-free medium. The cells were harvested following incubating the cells for 30 min in the absence or presence of 100 ng/ml CCN2, and nuclear extracts were prepared. Western blot analysis shows that CCN2 induces IκB phosphorylation and degradation (A) and translocation of the NF-κB subunits p65/p50 from cytoplasm to nucleus (B).
CCN2 promotes NF-κB DNA binding activity
To further explore whether active NF-κB can bind to its target DNA sequence and activate gene transcription in response to CCN2 stimulation, NF-κB DNA binding activity was determined by EMSA following incubation of HSC nuclear protein extracts with 32P-labeled NF-κB oligomers containing NF-κB/CBF1 binding sites. Two complexes were significantly enhanced by CCN2, reaching a plateau about 30 min after CCN2 addition (Fig. 2A). The NF-κB inhibitor, Bay11-7082, prevented complex formation when added to the cells prior to CCN2 treatment but not when added after CCN2 treatment (Figure 2B). As shown in Figure 2C, active NF-κB induced by CCN2 comprised three separate dimers (p65/p65, p65/p50 and p50/p50) based on the fact that a supershift (S1) was obtained with anti-p65 and anti-p50 antibodies with the concomitant disappearance of all three bands and that a supershift (S2) with anti-p65 antibody was associated with loss of the top two bands. Three dimers of NF-κB induced by CCN2 were also demonstrated following incubation of the nuclear extract with a CBF-1 mutant oligonucleotide probe but not with a p50/p65 mutant oligonucleotide probe (Figure 2D). These results indicate that CCN2 induces activation of NF-κB and its assembly into three dimers that individually bind to NF-κB DNA binding site.
Figure 2 Modulation of NF-κB DNA binding activity by CCN2. HSC were harvested at the desired time points after treatment with or without 100 ng/ml CCN2. 6 μg nuclear protein extract were used in 20 μl reactions, containing 0.2 ng 32P-labeled double strand NF-κB oligonucleotides. Reactions were fractioned through a nondenaturing 4% polyacrylamide gel. (A) Complex 1 (p65/p50) and complex 2 were enhanced after stimulation with CCN2. "Ctrl" represents a reaction lacking nuclear extract. (B) Bay11-7082 inhibited complex formation when added prior to CCN2 treatment ("Bay 11 + CCN2") but not when added subsequent to a 1 hour pretreatment with CCN2 ("CCN2 + Bay11"). (C) A supershift assay was performed by incubating pre-assembled gel shift assay complexes containing 8 μg nuclear extract with either 2 μg normal rabbit IgG or 2 μg anti-NF-κB antibody prior to separation through 8% polyacrylamide gel, showing that CCN2 stimulates the formation of an anti-p65/p65/anti-p50/p50/NF-κB oligonucleotide (S1) and an anti-p65/p65/NF-κB oligonucleotide (S2) supershift complexes. (D) A gel shift assay was performed following pre-incubating the nuclear extracts with either 32P-labeled p50/p65 site mutant oligonucleotides or CBF-1 site mutant oligonucleotides.
Effect of CCN2 on expression of NF-κB target genes
To explore the effect of CCN2 on transcriptional signaling by NF-κB, day 4 primary HSC cultures were transfected with luciferase reporter genes driven by either a minimal promoter alone (pTA) or together with four tandem copies of the NF-κB DNA binding sites (pNF-κB) that were identical to those used for the EMSA studies. pNF-κB generated 1.5-fold higher luciferase activities than pTA in HSC, whereas, the luciferase activities of pNF-κB in the cells following stimulation with CCN2 were 13-fold higher than control cells (Figure 3A). CCN2-mediated elevation of pNF-κB luciferase activity was completely abrogated by treatment of the cells with Bay11-7082, a specific inhibitor of IκBα phosphorylation (Figure 3B), suggesting that CCN2 modulates the transcriptional and translational event of NF-κB target genes via NF-κB signaling pathway.
Figure 3 Effect of CCN2 on expression of NF-κB response genes. (A) Freshly isolated rat HSC were placed in 12-well plates, transfected with 1 μg pTA-Luc or pNF-κB-Luc, and then incubated for another 24 h in the absence or presence of 100 ng/ml CCN2. (B) Transfected cells were pre-treated with 10 μM Bay11-7082 for 30 min and cultured for another 24 h in the absence or presence of 100 ng/ml CCN2. **P < 0.01 vs. pNF-κB-Luc; ##P < 0.01 vs. pNF-κB-Luc + CCN2.
CCN2 sustains HSC survival through NF-κB signaling pathway
To examine the effect of CCN2 on the fate of HSC, day 4-primary HSC cultures were treated with or without CCN2 for 24 h. Cell viability was determined by Trypan blue exclusion or by luminescent assessment of cellular ATP levels. As shown in Figures 4A and 4B, each assay showed that HSC viability was significantly elevated by CCN2. To determine if the NF-κB pathway was involved in this effect, Bay11-7082 was added to HSC cultures. As mentioned above, when added prior to CCN2, the inhibitor completely blocked the ability of CCN2 to stimulate NF-κB DNA binding activity whereas it had little inhibitory effect on DNA binding activity in cells that has been pre-treated with CCN2 (Figure 2B). While Bay11-7082 inhibited cell survival as expected, it had little inhibitory effect in cells that had been pre-treated with CCN2 (Figure 4A,B), consistent with the data shown in Figure 2B and supportive of the notion that prior stimulation of NF-κB by CCN2 was sufficient to overcome the effects of subsequently blocking NF-κB with Bay 11. Similarly, Bay-11-induced caspase-3 activity in HSC was reduced as much as 25% in CCN2-stimulated cells, consistent with the ability of CCN2 to rescue cells from apoptosis (Figure 4C).
Figure 4 Effect of CCN2 in sustaining HSC survival. Freshly isolated rat HSC were cultured in 6-well plates in 5% FBS DMEM for 24 h, followed by serum deprivation for 48 h. The cells were cultured for another 24 h in the absence or presence of 100 ng/ml CCN2 ("CCN2"). In the CCN2 protection assay, the cells were incubated with 10 μM Bay11-7082 for 24 h alone ("Bay11") or following pre-treatment of the cells with CCN2 for 1 h ("CCN2+Bay11"). For the Bay11 blocking assay, the cells were pre-treated with Bay11-7082 for 30 min and cultured for another 24 h in the presence of 100 ng/ml CCN2 ("Bay11+CCN2"). At the end of incubation time period, (A) cells were trypsinized and survival was determined by Trypan blue exclusion, or (B) cell viability was also quantified by measurement of the fluorescence intensity using CellTiter-Glo™ reagent, or (C) cell apoptosis was assessed by measurement of caspase-3 activity at 405 nm using a luminescence assay kit. *P < 0.05 vs. control; #P < 0.05 vs. CCN2 group; **P < 0.01 and * P < 0.05 vs. Bay11-7082 group.
Collectively, these data suggest that CCN2 is a survival factor for HSC and that survival is regulated via NF-κB signaling.
Discussion
CCN2 has emerged as a key mediator of fibrosis in both acute and chronic diseases [8-12]. In the liver, CCN2 expression is associated with hepatic fibrosis in both human subjects and animal models [10-12,14,17,31]. CCN2 appears to be directly involved in HSC biology as it is produced as a function of activation or exposure the cells to various fibrogenic stimuli including transforming growth factor-β, platelet-derived growth factor, alcohol and acetaldehyde [14]. Additionally, CCN2 promotes HSC adhesion, migration, proliferation, and synthesis of collagen type I [14,19], all of which are properties of activated HSC. [1-4,32]. Activation of HSC is regulated by several soluble factors, including growth factors, cytokines, and products of oxidative stress, as well as by extensive changes in the composition and organization of the extracellular matrix. HSC activation has previously been linked to activation of NF-κB while over-expression of IκBα in HSC has been shown to suppress NF-κB activation [26-29]. Our data show that serum-starved day 4 HSC demonstrate very low levels of nuclear NF-κB and no detectable DNA-binding activity. Following CCN2 stimulation, a marked nuclear translocation of NF-κB was evident along with a persistent DNA binding activity of its three dimers, and an induction of IκBα phosphorylation and degradation. Collectively, these data show that CCN2 activates the NF-κB signaling pathway in HSC. Moreover, as assessed using a NF-κB reporter construct, we showed that NF-κB response gene expression is induced by CCN2 in day 4 cultured HSC and support previous findings that NF-κB response genes are induced in activated HSC [26,29].
Apoptosis has been described as the nexus between liver injury and fibrosis [33] and increasing evidence suggests that NF-κB is involved in survival pathways in multiple hepatic cell types. For example, NF-κB prevents hepatocyte apoptosis following liver regeneration or exposure to tumor necrosis factor-α [34-36]. In addition, NF-κB also protects hepatocarcinoma cells or activated HSC from apoptosis [26,29,37]. Consistent with this latter observation, we showed that CCN2 promoted HSC survival and that NF-κB was involved in the response as shown by the finding that pretreatment of HSC with CCN2 protected the cells from Bay11-7082-induced decreased cell survival and increased caspase-3 activity.
The ability of CCN2 to sustain HSC survival supports previous observations showing that CCN2 is a survival factor for other cell types such as endothelial cells or chicken embryo fibroblasts [38,39]. Additionally, the related molecule, CCN1 (also known as CYR61) promotes anti-apoptotic pathways when over-expressed in breast cancer MCF7 cells in an integrin-dependent manner [40], consistent with the recognition of integrins as signaling receptors for CCN proteins [41]. In the case of HSC, the principal CCN2 receptor is integrin αvβ3 [19] which transduces survival signals in activated HSC [30] as well as during angiogenesis, wound healing, osteoporosis, and tumor metastasis [42-51]. Apoptosis in HSC is inhibited by engagement of integrin αvβ3 [30] and disruption of integrin-mediated HSC adhesion leads to induction of apoptosis [52]. Since expression of CCN2 and integrins is enhanced during HSC activation and liver fibrosis [53-57], persistence of the activated fibrogenic phenotype in HSC may occur, at least in part, by NF-κB survival pathways that are triggered via the binding of CCN2 to its integrin αvβ3 receptor.
Activation of HSC is also associated with the expression of death receptors such as Fas and TRAIL-R2, suggesting that HSC fate is likely determined by a balance between survival and apoptotic stimuli [33]. For example, HSC undergo apoptosis following treatment with nerve growth factor, a response that is due to the expression of the p75 nerve growth factor receptor [58] which has recently been implicated as a CCN2 signaling molecule in kidney mesangial cells [59]. Furthermore, pro-apoptotic effects of CCN2 have been reported in vascular smooth muscle cells and breast cancer cells, although the underlying mechanisms have yet to be understood [60-62]. Thus, depending on the presence and activity of its cognate cell surface receptors and their associated signaling pathways, CCN2 may be able to drive either apoptosis or survival in HSC. This points to a complex scenario whereby CCN2 may exert apparently opposing or contradictory effects on HSC viability, and future investigations will need to clarify this issue. Nonetheless, clinical fibrosis is now regarded as a largely reversible process that is strongly linked to apoptosis of activated HSC [33,63] and our data showing that CCN2 can promote HSC survival via NF-κB provide support for the development of new anti-fibrotic strategies that target CCN2, its receptors, or its signaling pathways.
Conclusion
In addition to promoting HSC fibrogenesis [17], CCN2 confers a survival advantage on HSC which is attributable, at least in part, to its ability to activate NF-κB signaling pathways in the cells.
Methods
Isolation and culture of HSC
In a protocol approved by the Institutional Animal Care and Use Committee of Children's Research Institute, Columbus, OH, primary HSC were isolated from normal male Sprague-Dawley rats as described [19]. Cells were grown in Dulbecco's modified Eagle's medium (DMEM; Gibco, Grand Island, NY, USA) supplemented with 10% fetal bovine serum (FBS), 100 U/ml penicillin and 100 μg/ml streptomycin. Cells were placed in 20 × 100 mm cell culture dishes (Falcon; Becton Dickinson, Franklin Lakes, NJ, USA) for nuclear extraction, 6-well tissue culture plates (Falcon) for Western blot and for cell viability assays, 12-well tissue culture plates (Falcon) for luciferase reporter gene transfection. The cells were then cultured DMEM/5% FBS for 24 h, followed by serum-free medium for 48 h. On day 4, the cells were treated with or without 100 ng/ml CCN2, and harvested at the desired time points. Human recombinant 38 kDa CCN2 was produced in a Chinese hamster ovary cell expression system as described [15].
Electrophoretic mobility shift assay (EMSA)
Nuclear extracts were prepared as described [64]. 32P-end-labeled double-stranded oligonucleotide probes used in this study comprised either wild type NF-κB oligonucleotide (sense: 5'-tgaggggactttcccagg-3'), p50/p65 mutant oligonucleotide (sense: 5'-tgaggcgactttcccagg-3') or CBF1-mutant oligonucleotide (sense: 5'-tgaggggacttcccgagg-3') [23]. The double-stranded NF-κB oligmers were used in nuclear protein-DNA binding reactions (20 μl volume) in which 1 μg poly dI:dC and 6 μg nuclear protein extract were incubated for 20 min at 4°C prior to addition of 0.2 ng 32P-labled double-stranded oligonucleotide for 30 min at 4°C. The contents of each tube were electrophoresed on non-denaturing 4% polyacrylamide gels which were then dried and analyzed by autoradiography. Supershift assays were performed by incubating pre-assembled gel shift assay complexes containing 8 μg nuclear extract with either 2 μg rabbit normal IgG, 2 μg rabbit polyclonal anti- p65 NF-κB IgG or/and 2 μg rabbit polyclonal anti- p50 NF-κB IgG (Santa Cruz Biotechnology Inc, CA, USA) for 2 h at 4°C before electrophoresis. The samples were then electrophoresed on 8 % polyacrylamide gels [65].
Transfections and luciferase assay
HSC were transfected with pNF-κB-Luc or pTA-Luc control vector using Superfect transfection reagent (QIAGEN, Valencia, CA, USA) under serum-free conditions for 3 h. The transfected cells were incubated for another 24 h in the absence or presence of 100 ng/ml CCN2. After normalization of transfection efficiency by β-galactosidase expression, luciferase enzyme activity was then quantified using a reporter assay kit (Clontech, Palo Alto, CA, USA).
SDS-PAGE and immunoblotting
25 μg cytoplasmic or nuclear extracts (see above) were subjected to SDS-PAGE in 5–15% gradient gels at 120 V for 1.5 h. Proteins were transferred to nitrocellulose membranes which were individually incubated with 1:500 dilutions of rabbit anti-IκBα, -phospho-IκBα, -NF-κB p65, or -NF-κB p50 polyclonal IgG (Santa Cruz Biotechnology Inc, CA, USA) in 5% nonfat milk TBST for 24 h at 4°C. The filters were then incubated with 1:1000 dilutions of HRP-conjugated goat anti-rabbit IgG for 1 h at room temperature. The membrane was washed extensively before detection using chemiluminescence.
Cell viability assay
Freshly isolated HSC were placed in 6-well tissue culture plates and incubated in DMEM/5% FBS for 24 h followed by serum-free medium for 48 h. The cells were incubated for an additional 24 h in the absence or presence of 100 ng/ml CCN2. In CCN2 protection assays, 10 μM Bay11-7082 was added to the medium either alone or following treatment of the cells with CCN2 for 1 h. At the end of the incubation period, the cells were trypsinized, mixed 1:1 with Trypan Blue solution (Sigma) and counted within 3 minutes under light microscopy using a hemocytometer.
In an alternative approach for measurement of cell viability, the CellTiter-Glo™ Luminescent assay kit was employed to assess the relative levels of cellular ATP. At the end of the incubation period, cells were treated with CellTiter-Glo™ reagent according to the manufacturer's instructions. Fluorescence was measured using black/clear tissue culture plates (BD Biosciences, Bedford, MA, USA). Cell viability was quantified by measurement of the sample fluorescence intensity at 560EX/590EM.
Caspase-3 activity assay
Freshly isolated HSC were placed in 12-well tissue culture plate, and incubated in DMEM/5% FBS for 24 h followed by serum-free medium for 48 h. The cells were incubated for an additional 24 h in the presence or absence of 100 ng/ml CCN2. 10 μM Bay11-7082 was added to the medium either alone or following treatment the cells with CCN2 for 1 h. Protein extracts were prepared following manufacturer's instructions. Caspase-3 activity was measured using an assay kit (Promega, Madison, WI, USA) in which cell extracts were mixed with Ac-DEVD-pNA substrate for 1-hour incubation at 30°C in 96-well microtiter plates prior to colorimetric measurement of p-nitroanilide product at 405 nm.
Statistical analysis
Data are presented as mean ± SE. Differences were analyzed statistically with paired sample student's t-test.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
RG carried out the experiments and wrote the draft manuscript. DRB participated in the experimental design and edited the manuscript.
Acknowledgements
This work was supported by NIH grant AA12817 awarded to D.R.B.
==== Refs
Gressner AM Bachem MG Cellular communications and cell-matrix interactions in the pathogenesis of fibroproliferative diseases: liver fibrosis as a paradigm Ann Biol Clin (Paris) 1994 52 205 226 7998677
Pinzani M Novel insights into the biology and physiology of the Ito cell Pharmacol Ther 1995 66 387 412 7667403 10.1016/0163-7258(94)00072-B
Eng FJ Friedman SL Fibrogenesis I. New insights into hepatic stellate cell activation: the simple becomes complex Am J Physiol Gastrointest Liver Physiol 2000 279 G7 G11 10898741
Friedman SL Molecular regulation of hepatic fibrosis, an integrated cellular response to tissue injury J Biol Chem 2000 275 2247 2250 10644669 10.1074/jbc.275.4.2247
Pinzani M Gesualdo L Sabbah GM Abboud HE Effects of platelet-derived growth factor and other polypeptide mitogens on DNA synthesis and growth of cultured rat liver fat-storing cells J Clin Invest 1989 84 1786 1793 2592560
Friedman SL Cytokines and fibrogenesis Semin Liver Dis 1999 19 129 140 10422196
Brigstock DR The connective tissue growth factor/cysteine-rich 61/nephroblastoma overexpressed (CCN) family Endocr Rev 1999 20 189 206 10204117 10.1210/er.20.2.189
Igarashi A Nashiro K Kikuchi K Sato S Ihn H Fujimoto M Grotendorst GR Takehara K Connective tissue growth factor gene expression in tissue sections from localized scleroderma, keloid, and other fibrotic skin disorders J Invest Dermatol 1996 106 729 733 8618012 10.1111/1523-1747.ep12345771
Lasky JA Ortiz LA Tonthat B Hoyle GW Corti M Athas G Lungarella G Brody A Friedman M Connective tissue growth factor mRNA expression is upregulated in bleomycin-induced lung fibrosis Am J Physiol 1998 275 L365 71 9700098
Abou-Shady M Friess H Zimmermann A di Mola FF Guo XZ Baer HU Buchler MW Connective tissue growth factor in human liver cirrhosis Liver 2000 20 296 304 10959808 10.1034/j.1600-0676.2000.020004296.x
Williams EJ Gaca MD Brigstock DR Arthur MJ Benyon RC Increased expression of connective tissue growth factor in fibrotic human liver and in activated hepatic stellate cells J Hepatol 2000 32 754 761 10845662 10.1016/S0168-8278(00)80244-5
Paradis V Dargere D Vidaud M De Gouville AC Huet S Martinez V Gauthier JM Ba N Sobesky R Ratziu V Bedossa P Expression of connective tissue growth factor in experimental rat and human liver fibrosis Hepatology 1999 30 968 976 10498649 10.1002/hep.510300425
Gao R Brigstock DR Low density lipoprotein receptor-related protein (LRP) is a heparin-dependent adhesion receptor for connective tissue growth factor (CTGF) in rat activated hepatic stellate cells Hepatol Res 2003 27 214 220 14585398 10.1016/S1386-6346(03)00241-9
Paradis V Dargere D Bonvoust F Vidaud M Segarini P Bedossa P Effects and regulation of connective tissue growth factor on hepatic stellate cells Lab Invest 2002 82 767 774 12065687
Ball DK Moussad EE Rageh MA Kemper SA Brigstock DR Establishment of a recombinant expression system for connective tissue growth factor (CTGF) that models CTGF processing in utero Reproduction 2003 125 271 284 12578541 10.1530/rep.0.1250271
Ball DK Rachfal AW Kemper SA Brigstock DR The heparin-binding 10 kDa fragment of connective tissue growth factor (CTGF) containing module 4 alone stimulates cell adhesion J Endocrinol 2003 176 R1 7 12553878 10.1677/joe.0.176R001
Rachfal AW Brigstock DR Connective tissue growth factor (CTGF/CCN2) in hepatic fibrosis Hepatol Res 2003 26 1 9 12787797 10.1016/S1386-6346(03)00115-3
Gao R Ball DK Perbal B Brigstock DR Connective tissue growth factor induces c-fos gene activation and cell proliferation through p44/42 MAP kinase in primary rat hepatic stellate cells J Hepatol 2004 40 431 438 15123357 10.1016/j.jhep.2003.11.012
Gao R Brigstock DR Connective tissue growth factor (CCN2) induces adhesion of rat activated hepatic stellate cells by binding of its C-terminal domain to integrin {alpha}v{beta}3 and heparan sulfate proteoglycan J Biol Chem 2004 279 8848 8855 14684735 10.1074/jbc.M313204200
Ghosh S May MJ Kopp EB NF-kappa B and Rel proteins: evolutionarily conserved mediators of immune responses Annu Rev Immunol 1998 16 225 260 9597130 10.1146/annurev.immunol.16.1.225
Ghosh S Karin M Missing pieces in the NF-kappaB puzzle Cell 2002 109 Suppl S81 96 11983155 10.1016/S0092-8674(02)00703-1
Berti R Williams AJ Moffett JR Hale SL Velarde LC Elliott PJ Yao C Dave JR Tortella FC Quantitative real-time RT-PCR analysis of inflammatory gene expression associated with ischemia-reperfusion brain injury J Cereb Blood Flow Metab 2002 22 1068 1079 12218412 10.1097/00004647-200209000-00004
Oakley F Mann J Ruddell RG Pickford J Weinmaster G Mann DA Basal expression of IkappaBalpha is controlled by the mammalian transcriptional repressor RBP-J (CBF1) and its activator Notch1 J Biol Chem 2003 278 24359 24370 12700242 10.1074/jbc.M211051200
Mann DA Smart DE Transcriptional regulation of hepatic stellate cell activation Gut 2002 50 891 896 12010897 10.1136/gut.50.6.891
Sonenshein GE Rel/NF-kappa B transcription factors and the control of apoptosis Semin Cancer Biol 1997 8 113 119 9299589 10.1006/scbi.1997.0062
Elsharkawy AM Wright MC Hay RT Arthur MJ Hughes T Bahr MJ Degitz K Mann DA Persistent activation of nuclear factor-kappaB in cultured rat hepatic stellate cells involves the induction of potentially novel Rel-like factors and prolonged changes in the expression of IkappaB family proteins Hepatology 1999 30 761 769 10462383 10.1002/hep.510300327
Hellerbrand C Jobin C Licato LL Sartor RB Brenner DA Cytokines induce NF-kappaB in activated but not in quiescent rat hepatic stellate cells Am J Physiol 1998 275 G269 78 9688654
Rippe RA Schrum LW Stefanovic B Solis-Herruzo JA Brenner DA NF-kappaB inhibits expression of the alpha1(I) collagen gene DNA Cell Biol 1999 18 751 761 10541434 10.1089/104454999314890
Lang A Schoonhoven R Tuvia S Brenner DA Rippe RA Nuclear factor kappaB in proliferation, activation, and apoptosis in rat hepatic stellate cells J Hepatol 2000 33 49 58 10905586 10.1016/S0168-8278(00)80159-2
Zhou X Murphy FR Gehdu N Zhang J Iredale JP Benyon RC Engagement of alpha Vbeta 3 integrin regulates proliferation and apoptosis of hepatic stellate cells J Biol Chem 2004
Hayashi N Kakimuma T Soma Y Grotendorst GR Tamaki K Harada M Igarashi A Connective tissue growth factor is directly related to liver fibrosis Hepatogastroenterology 2002 49 133 135 11941937
Gaca MD Zhou X Benyon RC Regulation of hepatic stellate cell proliferation and collagen synthesis by proteinase-activated receptors J Hepatol 2002 36 362 369 11867180 10.1016/S0168-8278(01)00285-9
Canbay A Friedman S Gores GJ Apoptosis: the nexus of liver injury and fibrosis Hepatology 2004 39 273 278 14767974 10.1002/hep.20051
Iimuro Y Nishiura T Hellerbrand C Behrns KE Schoonhoven R Grisham JW Brenner DA NFkappaB prevents apoptosis and liver dysfunction during liver regeneration J Clin Invest 1998 101 802 811 9466975
Black D Bird MA Hayden M Schrum LW Lange P Samson C Hatano E Rippe RA Brenner DA Behrns KE TNF alpha-induced hepatocyte apoptosis is associated with alterations of the cell cycle and decreased stem loop binding protein Surgery 2004 135 619 628 15179368 10.1016/j.surg.2003.11.004
Imose M Nagaki M Naiki T Osawa Y Brenner DA Asano T Hayashi H Kato T Moriwaki H Inhibition of nuclear factor kappaB and phosphatidylinositol 3-kinase/Akt is essential for massive hepatocyte apoptosis induced by tumor necrosis factor alpha in mice Liver Int 2003 23 386 396 14708901 10.1034/j.1478-3231.2003.00867.x
Chiao PJ Na R Niu J Sclabas GM Dong Q Curley SA Role of Rel/NF-kappaB transcription factors in apoptosis of human hepatocellular carcinoma cells Cancer 2002 95 1696 1705 12365017 10.1002/cncr.10829
Babic AM Chen CC Lau LF Fisp12/mouse connective tissue growth factor mediates endothelial cell adhesion and migration through integrin alphavbeta3, promotes endothelial cell survival, and induces angiogenesis in vivo Mol Cell Biol 1999 19 2958 2966 10082563
Gygi D Zumstein P Grossenbacher D Altwegg L Luscher TF Gehring H Human connective tissue growth factor expressed in Escherichia coli is a non-mitogenic inhibitor of apoptosis Biochem Biophys Res Commun 2003 311 685 690 14623326 10.1016/j.bbrc.2003.10.061
Lin MT Chang CC Chen ST Chang HL Su JL Chau YP Kuo ML Cyr61 expression confers resistance to apoptosis in breast cancer MCF-7 cells by a mechanism of NF-kappaB-dependent XIAP up-regulation J Biol Chem 2004 279 24015 24023 15044484 10.1074/jbc.M402305200
Lau LF Lam SC The CCN family of angiogenic regulators: the integrin connection Exp Cell Res 1999 248 44 57 10094812 10.1006/excr.1999.4456
Cheresh DA Spiro RC Biosynthetic and functional properties of an Arg-Gly-Asp-directed receptor involved in human melanoma cell attachment to vitronectin, fibrinogen, and von Willebrand factor J Biol Chem 1987 262 17703 17711 2447074
Cheresh DA Pytela R Pierschbacher MD Klier FG Ruoslahti E Reisfeld RA An Arg-Gly-Asp-directed receptor on the surface of human melanoma cells exists in an divalent cation-dependent functional complex with the disialoganglioside GD2 J Cell Biol 1987 105 1163 1173 2443507 10.1083/jcb.105.3.1163
Gladson CL Cheresh DA Glioblastoma expression of vitronectin and the alpha v beta 3 integrin. Adhesion mechanism for transformed glial cells J Clin Invest 1991 88 1924 1932 1721625
Schwartz MA Signaling by integrins: implications for tumorigenesis Cancer Res 1993 53 1503 1506 8453614
Brooks PC Clark RA Cheresh DA Requirement of vascular integrin alpha v beta 3 for angiogenesis Science 1994 264 569 571 7512751
Brooks PC Montgomery AM Rosenfeld M Reisfeld RA Hu T Klier G Cheresh DA Integrin alpha v beta 3 antagonists promote tumor regression by inducing apoptosis of angiogenic blood vessels Cell 1994 79 1157 1164 7528107 10.1016/0092-8674(94)90007-8
Montgomery AM Reisfeld RA Cheresh DA Integrin alpha v beta 3 rescues melanoma cells from apoptosis in three-dimensional dermal collagen Proc Natl Acad Sci U S A 1994 91 8856 8860 7522323
Drake CJ Cheresh DA Little CD An antagonist of integrin alpha v beta 3 prevents maturation of blood vessels during embryonic neovascularization J Cell Sci 1995 108 ( Pt 7) 2655 2661 7593306
Scatena M Almeida M Chaisson ML Fausto N Nicosia RF Giachelli CM NF-kappaB mediates alphavbeta3 integrin-induced endothelial cell survival J Cell Biol 1998 141 1083 1093 9585425 10.1083/jcb.141.4.1083
Miyauchi A Alvarez J Greenfield EM Teti A Grano M Colucci S Zambonin-Zallone A Ross FP Teitelbaum SL Cheresh D Recognition of osteopontin and related peptides by an alpha v beta 3 integrin stimulates immediate cell signals in osteoclasts J Biol Chem 1991 266 20369 20374 1939092
Iwamoto H Sakai H Tada S Nakamuta M Nawata H Induction of apoptosis in rat hepatic stellate cells by disruption of integrin-mediated cell adhesion J Lab Clin Med 1999 134 83 89 10402063 10.1016/S0022-2143(99)90057-4
Reeves HLDCLDCPBAD Wisse EKDLBC ß1-integrin expression in hepatic stellate cells from normal and diseased human livers: 1997; Bordeaux, France,.
1997 6 The Kupffer Cell Foundation, Leiden, The Netherlands 154 155
Richter HB Franke H Dargel R Expression of tenascin, fibronectin, and laminin in rat liver fibrogenesis--a comparative immunohistochemical study with two models of liver injury Exp Toxicol Pathol 1998 50 315 322 9784003
Carloni V Romanelli RG Pinzani M Laffi G Gentilini P Expression and function of integrin receptors for collagen and laminin in cultured human hepatic stellate cells Gastroenterology 1996 110 1127 1136 8613002 10.1053/gast.1996.v110.pm8613002
Iwamoto H Sakai H Nawata H Inhibition of integrin signaling with Arg-Gly-Asp motifs in rat hepatic stellate cells J Hepatol 1998 29 752 759 9833913 10.1016/S0168-8278(98)80256-0
Kato R Kamiya S Ueki M Yajima H Ishii T Nakamura H Katayama T Fukai F The fibronectin-derived antiadhesive peptides suppress the myofibroblastic conversion of rat hepatic stellate cells Exp Cell Res 2001 265 54 63 11281643 10.1006/excr.2001.5179
Trim N Morgan S Evans M Issa R Fine D Afford S Wilkins B Iredale J Hepatic stellate cells express the low affinity nerve growth factor receptor p75 and undergo apoptosis in response to nerve growth factor stimulation Am J Pathol 2000 156 1235 1243 10751349
Wahab NA Weston BS Mason RM Connective tissue growth factor CCN2 interacts with and activates the tyrosine kinase receptor TrkA J Am Soc Nephrol 2005 16 340 351 15601748 10.1681/ASN.2003100905
Hishikawa K Nakaki T Fujii T Connective tissue growth factor induces apoptosis via caspase 3 in cultured human aortic smooth muscle cells Eur J Pharmacol 2000 392 19 22 10748267 10.1016/S0014-2999(00)00115-1
Hishikawa K Oemar BS Tanner FC Nakaki T Fujii T Luscher TF Overexpression of connective tissue growth factor gene induces apoptosis in human aortic smooth muscle cells Circulation 1999 100 2108 2112 10562268
Hishikawa K Oemar BS Tanner FC Nakaki T Luscher TF Fujii T Connective tissue growth factor induces apoptosis in human breast cancer cell line MCF-7 J Biol Chem 1999 274 37461 37466 10601320 10.1074/jbc.274.52.37461
Friedman SL Arthur MJ Reversing Hepatic Fibrosis Science & Medicine 2002 194 205
Dignam JD Lebovitz RM Roeder RG Accurate transcription initiation by RNA polymerase II in a soluble extract from isolated mammalian nuclei Nucleic Acids Res 1983 11 1475 1489 6828386
Pizzi M Goffi F Boroni F Benarese M Perkins SE Liou HC Spano P Opposing roles for NF-kappa B/Rel factors p65 and c-Rel in the modulation of neuron survival elicited by glutamate and interleukin-1beta J Biol Chem 2002 277 20717 20723 11912207 10.1074/jbc.M201014200
|
16303051
|
PMC1308830
|
CC BY
|
2021-01-04 16:39:15
|
no
|
Cell Commun Signal. 2005 Nov 22; 3:14
|
utf-8
|
Cell Commun Signal
| 2,005 |
10.1186/1478-811X-3-14
|
oa_comm
|
==== Front
Cerebrospinal Fluid ResCerebrospinal Fluid Research1743-8454BioMed Central London 1743-8454-2-101629318810.1186/1743-8454-2-10ResearchMemory and selective learning in children with spina bifida-myelomeningocele and shunted hydrocephalus: A preliminary study Vachha Behroze [email protected] Richard C [email protected] Pediatric Developmental Disabilities, Texas Scottish Rite Hospital for Children, Dallas, TX, 75219, USA2 Department of Pediatrics, University of Texas Southwestern Medical Center, Dallas, TX, 75235, USA2005 17 11 2005 2 10 10 21 7 2005 17 11 2005 Copyright © 2005 Vachha and Adams; licensee BioMed Central Ltd.2005Vachha and Adams; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Selective learning is the ability to select items of relevance from among less important items. Limited evidence exists regarding the efficiency with which children with spina bifida-myelomeningocele and shunted hydrocephalus (SB/SH) are able to learn information. This report describes initial data related to components of learning and metacognitive skills in children with SB/SH.
Methods
Twenty six children with SB/SH and 26 controls (age: 7 – 16 y) with average intelligence, and monolingual English-speaking backgrounds participated in the study. Exclusion criteria for the SB/SH group were: prior history of shunt infection, history of seizure or shunt malfunction within the previous three months, prior diagnoses of attention disorders and/or clinical depression. Children were presented lists of words with equal exemplars each of two distinct semantic categories (e.g. fruits, animals), and told to make as high a score as possible by learning the words. The value of the words was designated by category membership (e.g. animals = low value; fruits = high value). The total number of words learned across three learning trials was used to determine memory span. Selective learning efficiency (SLE) was computed as the efficiency with which items of greater value were selectively learned across three trials.
Results
Children with SB/SH did worse than controls on memory span (P < 0.05). Although SLE was not significantly different between groups, when asked what strategy was used in the selective learning tasks, 65% of the SB/SH children said they tried to remember all words (inefficient strategy). In contrast, 85% of controls said they tried to remember the higher value words – the more efficient strategy.
Conclusion
Success in school is often dependent on the ability to recall important facts selectively and ignore less important information. Children with SB/SH in our study had a poor memory span and were unable to monitor and report an efficient and workable metacognitive strategy required to remember a list of words. Preliminary findings may begin to explain our previous clinical and research findings wherein children with SB/SH often focus on extraneous details, but demonstrate difficulty remembering the main gist of a story/event.
==== Body
Background
The potential barriers capable of impacting daily learning, recall, and functional application of prior knowledge (in rehearsed or novel situations) among children with spina bifida-myelomeningocele and shunted hydrocephalus (SB/SH) are multifaceted and tightly inter-related. Neurological insults include ventriculomegaly of varying degrees, callosal formation of varying completion, Chiari II malformation with varying complexity, and others [1]. By their nature, these neurological differences occur along a continuum rather than a simple, discrete, all-or-none classification.
Variations in cognitive functions impacting learning among children with SB/SH are similarly complex. The overwhelming majority of such youngsters have average intelligence based on standardized tests [2]. Prior studies have described potential barriers to learning in this population based on differences in language abilities and in temperament characteristics [3,4]. Given the underlying developmental differences in brain structure, other cognitive skills such as memory, as well as metacognitive skills, might be expected to contribute to the learning difficulties demonstrated by these children.
We report here a preliminary study of memory and metacognitive skills, specifically 'selective learning' abilities, in children with SB/SH and their impact on learning. Selective learning (SL) has been defined as the ability to select and learn particular items of higher value from a broader array of available information [5]. This process has been described in typically developing children (6 to 16 years) and among adults and children with post-natal traumatic brain injury (TBI) of various etiologies [5-8]. Among individuals with TBI, a better understanding of how the child subsequently selects, focuses on, and learns salient facts may be central to success both within the classroom and within home / community settings. In each of these environments, everyday circumstances require quick and appropriate responses.
For children with SB/SH, brain abnormalities begin in the first month of gestation and are sustained beyond birth [1,9-11]. These developmental disabilities are fundamentally different from those of children with previously typical development who subsequently sustain brain injury. Nevertheless, their ability to learn salient information selectively is at least as critical for successful performance both in academic and life skills. Since memory capacity limits the ability to recall all information encountered, successful classroom learning is dependent, in part, on the child's ability to selectively learn and recall the important facts in comparison to less important information [12]. Studying selective learning and metacognitive learning strategies employed by children with SB/SH should provide professionals (rehabilitation and school teachers, alike) important insights into an individual child's specific strengths and weaknesses so that strategies/methods of teaching can be optimized.
The primary aim of this study was to investigate the ability of children with SB/SH to learn selectively items based on differential value in comparison to typically developing peers. Additionally, in order to determine if there were dissociations within the same task between the total words recalled without regard to value (memory span) and the more complex processes involved in selective learning efficiency, we investigated differences in the memory span of children with SB/SH and their controls. Finally, we computed and compared the number of children in each group that reported the most efficient strategy that might help them learn and remember the items in order to be most successful in a learning task.
Methods
Participants
Twenty-six children with SB/SH between the ages of 7–16 years (M = 12.3y, SD = 2.7) were recruited from the Spina Bifida Program at Texas Scottish Rite Hospital for Children (TSRHC), Dallas, Texas. Twenty-six typically developing children from the same geographic region, aged 7–16 years (M = 11.2y, SD = 2.6) served as controls. Research reported in the manuscript was performed with the approval of the ethics committee of the Institutional Review Board at the Texas Scottish Rite Hospital for Children and University of Texas Southwestern Medical Center, Dallas and was in compliance with the Helsinki declaration (Permission Number: 0603-421).
To minimize confounding variables, strict inclusion and exclusion criteria were created and followed. Specifically, inclusion criteria for the SB/SH group included: 1) Diagnosis of myelomeningocele with Chiari II; 2) Hydrocephalus diagnosed, and VP shunt placed within the first year of life. Inclusion criteria for both the SB/SH and control groups included: 1) English as primary language; 2) No history of seizures within the past three months. 3) Average intelligence (Full Scale IQ greater than 80 on the Wechsler Intelligence Scale for Children – Fourth Edition). Exclusion criteria for the SB/SH group were: 1) History of shunt infection; 2) Shunt malfunction with resultant symptoms within the previous three months. Exclusionary criteria for both groups included: 1) Uncorrected sensory (auditory or visual) acuity deficits or motor disabilities that would preclude a pointing response; 2) Prior diagnosis of mental retardation; 3) Prior diagnosis of attention deficit disorder.
Procedure
Children were presented lists of 14 spoken words comprising seven exemplars each of two distinct semantic categories (e.g. fruits and animals). Category exemplars for both categories comprised concrete, low age of acquisition (AoA) words drawn from the University of Western Australia Psycholinguistics Database [13], from a set of 297 object names normed by age of acquisition reported by Morrison et al. [14] and from samples of learning materials prescribed for kindergarteners in the Texas public education system [15]. AoA rating for words was set for 22 – 60 months of age. To further ensure that children in this study would have been exposed to these exemplars by age 5 years in the Texas public education system, selected words were rated for age of acquisition by a panel of three educators in the Texas public education system, each with at least six years of kindergarten to first grade teaching experience. If a discrepancy in concreteness or AoA rating for a word could not be resolved between the three educators, that word was discarded (for e.g. bear, bat having two meanings were removed from the original listing), yielding the final set of words deemed to be acquired by a majority of children by 5 years of age. Categories were counterbalanced such that if an exemplar in one category had a particular age of acquisition, then an exemplar with a similar age of acquisition was selected for the other category.
Each child was individually tested, and the list was spoken by a single examiner at the rate of one word per second. There was one practice trial and a total of three consecutive experimental trials presented in the same session.
Each child was told to make as high a score as possible by remembering the words. The value of the words was designated by category membership (e.g. name of any fruit = 10 points; name of any animal = 1 point). For each list, the value assigned to the categories was told to the child both prior to list presentation and after list presentation, immediately before recall, in order to eliminate memory for category-value designations as a confounding variable. The words were mixed in each list so that exemplars for the two categories occurred in random order within the list.
To ensure that the children were equally familiar with exemplars in both categories, each child was given a post-test screening task in which he/she was asked to: 1) name the category exemplars used in the experiment (shown to them as colored pictures); and 2) sort the category exemplars used in the experiment into the appropriate categories.
Colored pictures representing exemplars in each category were mixed randomly and presented in the same sequence to each child. One picture was presented at a time in the center of a computer screen. Participants were instructed to name each picture as quickly but as accurately as possible. Each picture remained on the screen until the child named it. As soon as the participant responded, the picture disappeared and a blank screen was presented for 1500 msec, followed by the next picture. Correct and incorrect responses were recorded manually on a record form. Response time was recorded in msec from onset of picture presentation to onset of response. The total number of correct responses for each category was computed and response times were calculated for correct responses only. If a child self-corrected an incorrect response, it was counted towards the total number of correct responses made by the child, but was excluded from response time calculation.
Given the low age of acquisition of the tests, children in both groups were able to name all category exemplars used in the experiment with 100% accuracy. SB/SH and control groups did not differ significantly in mean response times for any category (P > 0.05). Within group analysis showed children with SB/SH did not demonstrated significant differences in mean response time for naming exemplars in either category (P > 0.05). Children in the control group also did not demonstrated significant differences in mean response time for naming exemplars in either category (P > 0.05).
Next, the same sequence was presented, but this time the child was asked to provide the category for the exemplar. Correct and incorrect responses were recorded manually. The total number of correct responses for each category was computed and response times were calculated as stated above. Children in both groups were able to sort the category exemplars used in the experiment into the appropriate categories with 100% accuracy. Between and within group analysis demonstrated no significant differences in mean response time for sorting category exemplars for the two groups (P > 0.05).
Scoring
There were three main measures of interest. The first was the degree of selective learning efficiency demonstrated by the child learning preferentially the words of higher value (e.g. fruits) in comparison to words of lower value (e.g. animals). To ensure that the measure be independent of the total number of words recalled (i.e. the child's memory span), we calculated the selective learning efficiency of each trial for each subject using the following formula as given by Hanten and colleagues [5] : Selective learning efficiency (SLE) = (actual score - chance score)/(ideal score - chance score), where the actual score = the sum of the values of the items recalled; the chance score = the score that is predicted in the absence of selectivity given the number of items recalled (in other words, if half the words recalled were high value and half were low value); ideal score = maximum score that could be achieved given the number of items recalled. The range of the SLE index extends from -1 to 1 with a score of 0 indicating chance performance.
The second measure of interest was the total number of words recalled, which reflects the memory span of the child. This measure was the average number of words recalled across trials without regard to the word value and included the total number of high and low value words averaged across three trials.
The third measure of interest examined whether children reported the most efficient strategy needed to learn words in order to make the highest score possible on the selective learning task. Children in both groups were asked what would be the most efficient strategy required to remember the words in order to make the highest score on the learning task. Responses were recorded and coded 'efficient' if the response was "Try to remember high value word (e.g. fruits) more than low value word (e.g. animals)." or similar variant. Responses were considered 'inefficient' if the response was "I tried to remember all words while you said them" or similar variant. The percentage of children reporting the efficient versus inefficient strategies for each group was calculated.
Results
Preliminary analyses showed no significant group differences in age (t [50] = -1.47, P = 0.148), mean parental education, as proxy for socioeconomic status, (t [50] = 0.64, P = 0.526); gender (χ2[1] = 0.080, P = 0.778); or ethnicity (χ2[2] = 3.476, P = 0.176). No significant correlations were noted between mean parental education and mean SLE (P = 0.93) or mean memory span (P = 0.58). No significant differences were noted in performance by males versus females on measures of mean SLE (t [50] = 0.55, P = 0.58) or mean memory span (t [50] = 0.29, P = 0.77).
SLE
The mean SLE, not adjusted for age, for the SB/SH and control groups is provided in Table 1. Although the children with SB/SH appeared to have lower mean SLE averaged across the three trials than did typically developing children, the group differences failed to reach significance (F [1,50] = 0.98, P = 0.327). Pearson correlations indicated that age-at-test of the child was positively related to mean SLE (r = 0.36, P = 0.01) such that as the child's age increased, the SLE increased. Therefore, mean SLE averaged across the three learning trials was analyzed using analysis of covariance with group (SB/SH or control) as the independent variable and age-at-test as covariate, at a Pvalue of 0.05. No significant differences were noted in the SLE/age relationship as a function of group (age*group interaction F [1,48] = 1.73, P = 0.20). After adjusting for age, the differences between groups remained non significant (F [1,49] = 2.81, P = 0.10). There was a significant effect of age (F [1,49] = 9.56, P = 0.003).
Table 1 Age, SLE index and memory span for children with SB/SH and controls. Data are means (SD).
Group Age (years) SLE index Memory span
SB/SH 12.3 (2.7) 0.29 (0.34) 6.9 (1.9)
Control 11.2 (2.6) 0.39 (0.32) 8.9 (2.0)
A repeated measures analysis of covariance was used to compare the two groups (SB/SH or control), and evaluate consistency across the three trials for SLE, using age as covariate. The repeated measures analysis of covariance using Wilk's Lambda found no significant differences across the three trials (F [2,48] = 1.58, P = 0.22), a non significant interaction of trial*age (F [2,48] = 1.14, P = 0.33), and a non significant interaction of trial*group (F [2,48] = 0.30, P = 0.74).
Memory Span
The mean memory span, not adjusted for age, for the SB/SH and control groups is provided in Table 1. A significant effect of group indicated children with SB/SH had a significantly lower mean memory span averaged across the three trials (F [1,50] = 14.11, P < 0.001). Group means, not corrected for age are reported in Table 1. Pearson correlations indicated that age-at-test of the child was positively related to mean memory span (r = 0.40, P = 0.003) such that as the child's age increased, the memory span also increased. Therefore mean memory span averaged across the three learning trials was analyzed using analysis of covariance with group (SB/SH or control) as the independent variable and age-at-test as covariate, at a P value of 0.05. No significant differences were noted in the memory span/age relationship as a function of group (age*group interaction F [1,48] = 0.05, P = 0.82). After adjusting for age, there remained a significant difference between the two groups (F [1,49] = 29.76, P << 0.001), with controls having a higher mean than SB/SH children.
The repeated measures analysis of covariance with age as covariate and using Wilk's Lambda found no significant differences across the three trials (F [2,48] = 0.29, P = 0.75), a non significant interaction of trial*age (F [2,48] = 0.29, P = 0.75), and a non significant interaction of trial*group (F [2,48] = 1.77, P = 0.18).
Metacognitive Strategy
The majority of children with SB/SH were unable to monitor and report the most efficient strategy they should use in order to learn the words in a fashion that would support their making the highest possible score on the task. The majority of them (65%) reported that they tried to remember all the words equally as they perceived that to be the strategy that would most likely yield a high score (inefficient strategy, Fig. 1). In contrast, most children in the control group (85%) said they tried to remember the fruits more than the animals because fruits were worth much more than the animals (efficient strategy, Fig. 1). Statistical analysis (two-tailed Fisher Exact Test) confirmed that children with SB/SH were unable to monitor and report the learning strategy most likely to yield the highest score on the selective learning task (two-tailed P < 0.001).
Figure 1 Efficient/inefficient strategy reported by children with SB/SH and controls.
Discussion
A finding of this preliminary study was that although children with SB/SH appeared to have lower SLE averaged across the three trials than did typically developing children (Table 1), group differences failed to reach significance. This result may be attributed to the relatively small sample size used in the study. However, a more likely explanation may be related to the poor memory span demonstrated by children with SB/SH.
Consistent with previous studies of learning and memory in children with SB/SH [16,17] children with SB/SH in this study demonstrated significantly impaired memory span. In other words, children with SB/SH remembered fewer words overall irrespective of the value (higher or lower) of the words, suggesting the lack of significance on SLE tasks may be related to a pervasive memory impairment in this group.
Our findings are directly opposed to those seen in recent studies in children with TBI which showed impaired SLE but preserved memory span [5,6]. Budson and Price [18] recently reviewed different memory systems and their dependence on different neuroanatomical structures and variations. They described working memory (a component of selective learning) as "the ability to temporarily maintain and manipulate information that one needs to keep in mind" and dependent on "a network of activity that includes subcortical structures as well as frontal and parietal cortical regions" [18]. Just as injury to the orbitofrontal circuits secondary to trauma may play a role in differentially affecting more than one memory system in TBI [5,6], developmental variations unique to SB/SH might be expected to impact multiple, but potentially different components of the central nervous system than those involved in children with TBI. These individually and collectively, may impact learning and subsequent academic performance.
In addition to poor memory span performance, the results of this study also suggested that children with SB/SH were impaired relative to typically developing children in their ability to report the most efficient strategy in learning items of differential value in order to attain the highest possible score on a learning task. Learning and memory skills are inextricably linked and are critical components for optimal academic performance. Overt memory deficits are likely to be noted in neuropsychological evaluations, and when these children encounter professionals/parents within medical, school or home settings. Less dramatic, but equally important variations in the child's perception and recognition of useful learning strategies can also contribute to functional impairments within these environments. For instance, school-based texts have salient elements interspersed with less relevant information included in the service of making the text more interesting to learners. Success in school is potentially dependent on the ability to selectively learn and recall important facts and ignore less important information [5]. Educational teaching strategies suggest that an effective intervention for fact learning is to point out the salient components to be learned and provide incentives for learning [19]. The majority of children with SB/SH in our study were unable to reason out and report the incentive inherent in the selective learning task utilized in this study (i.e. more points for high value words than low value words). These preliminary findings are consistent with our previous clinical and research findings wherein children with SB/SH focus on many extraneous details, but are unable to remember the salient points of a story/event [20,21]. The results suggest a difficulty with metacognitive skills or the ability to think about appropriate learning strategies. Automatic strategies such as trying to remember as much information as possible regardless of value, taken together with the fact that these children already have an impaired memory span may account for the less than optimal academic performance described in these youngsters and warrants further investigation.
To ensure familiarity with exemplars in each category, children were presented with lists of words of low age of acquisition and concreteness. Post-test screening also revealed children in both groups were able to rapidly name and sort test items into the correct category with 100% accuracy. The category designated as high value was not, however, counterbalanced by subject in this study. While unlikely, it is possible that an individual child's preference of animals over fruits (or vice versa) could conceivably have resulted in a tendency bearing on his selective learning performance. This issue warrants consideration in future studies. Future studies that determine differences in selective learning performance based on whether words were presented auditorily or visually or both will help elucidate the types of memory and cognitive processes affected in children with SB/SH.
In summary, mood, anxiety, language and its various subsystem components, cognition (attention, abstract reasoning), temperament, family and environmental variables, and underlying autonomic instability have all been described among children with SB/SH [3,4,20,22]. Each of the above contributes to or detracts from the process of 'learning' whether in the classroom, or in the community. The concept of a 'spina bifida-hydrocephalus learning spectrum' might be more functional to parents, students, and educators than attempts to identify a single profile. To this spectrum, the various subsystems of memory and metacognition must be added. Further studies in each of these component arenas are likely to underscore the complexity of the child with SB/SH, but should allow the informed professional to bring useful information to assist the child in achieving his or her maximum learning potential.
List of abbreviations
SB/SH, shunted hydrocephalus and spina bifida-myelomeningocele; SLE, selective learning efficiency; AoA, age of acquisition
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Both authors contributed equally to this work. All authors read and approved the final manuscript.
Acknowledgements
The authors acknowledge the contributions of Dr. Richard Browne for statistical consultations, Kathi Allman, Beth Douek and Paula Craig-Williams for assistance with categorical exemplars selection, and Monte Davenport for support of this study.
==== Refs
Barkovich A Barkovich A Congenital malformations of the brain and skull Pediatric Neuroimaging 2000 Third Philadelphia, Lippincott Williams and Wilkins 251 383
Friedrich WN Lovejoy MC Shaffer J Shurtleff DB Beilke RL Cognitive abilities and achievement status of children with myelomeningocele: A contemporary sample. Journal of Pediatric Psychology 1991 16 423 428 1941424
Vachha B Adams R Language differences in young children with myelomeningocele and shunted hydrocephalus Pediatric Neurosurgery 2003 39 184 190 12944698 10.1159/000072469
Vachha B Adams R Myelomeningocele, Temperament Patterns, and Parental Perceptions. Pediatrics 2005 115 58 63
Hanten G Zhang L Levin H Selective learning in children after traumatic brain injury: a preliminary study Child Neuropsychology 2002 8 107 120 12638064
Hanten G Chapman SB Gamino JF Zhang L Benton SB Stallings-Roberson G Hunter JV Levin HS Verbal selective learning after traumatic brain injury in children. Annals of Neurology 2004 56 847 853 15562406 10.1002/ana.20298
O'Sullivan JT Preschooler's beliefs about effort, incentives and recall Journal of Experimental Child Psychology 1993 55 396 414 10.1006/jecp.1993.1022
Bauer R Peller-Porth V The effect of increased incentive on free recall by learning disabled and nondisabled children Journal of General Psychology 1991 117 447 461 2286838
Adams R Spina Bifida: Life Span Management. 2000 , American Physical Therapy Association
Miyan JSCDC Humanity Lost: The cost of Cortical Maldevelopment. Is There Light Ahead? European Journal of Pediatric Surgery 2001 11 54 59 10.1055/s-2001-19740
Vachha B Adams R Rollins N Depiction of limbic tract anomalies with diffusion tensor imaging and fiber tract reconstruction: Initial investigation of association with memory and learning in children with myelomeningocele and chiari II Radiology
Ward-Lonergan JM Liles BZ Anderson AM Listening comprehension and recall abilities in adolescents with language learning disabilities and without disabilities for social studies lectures Journal of Communication Disorders 1999 31 1 32 9421765 10.1016/S0021-9924(97)00048-8
MRC Psycholinguistic Database: Machine Usable Dictionary. Version 2.00
Morrison C Chappell T Ellis AW Age of acquisiton norms for a large set of object names adn their relation to adult estimates and other variables The Quarterly Journal of Experimental Psychology 1997 50A 528 559
Texas Essential Knowledge and Skills by Grade Level (Elementary), s.v. "TEXAS EDUCATION AGENCY
Vachha B Adams R Learning and Memory Differences in Children with Myelomeningocele and Shunted Hydrocephalus (Abstract). Developmental Medicine and Child Neurology 2004 46 16 17
Yeates KOEBLNBEDDC Verbal learning and memory in children with myelomeningocele Journal of Pediatric Psychology 1995 20 801 812 8558379
Budson AE Price BH Current concepts: memory dysfunction New England Journal of Medicine 2005 352 692 699 15716563 10.1056/NEJMra041071
Kurtz B Denhiere G and Rossi J Cognitive and metacognitive aspects of text processing Text and text processing 1991 North Holland, Elsevier Science 77 103
Dennis MJBBM The content of narrative discourse in children and adolescents after early-onset hydrocephalus and in normally developing age peers Brain Lang 1994 46 8131040
Vachha B Adams R Language sample analysis in children with myelomeningocele and shunted hydrocephalus European Journal of Pediatric Surgery 2003 13 36 37
Riddle R Morton A Sampson J Vachha B Adams R Performance on the NEPSY among children with spina bifida Archives of Clinical Neuropsychology 2004 20 243 248 15708733 10.1016/j.acn.2004.07.004
|
16293188
|
PMC1308831
|
CC BY
|
2021-01-04 16:37:38
|
no
|
Cerebrospinal Fluid Res. 2005 Nov 17; 2:10
|
utf-8
|
Cerebrospinal Fluid Res
| 2,005 |
10.1186/1743-8454-2-10
|
oa_comm
|
==== Front
Chiropr OsteopatChiropractic & Osteopathy1746-1340BioMed Central London 1746-1340-13-241630768710.1186/1746-1340-13-24DebateA Chiropracticness Test Charlton Keith H [email protected] School of Medicine, Mayne Medical School, University of Queensland, Herston, Queensland 4006, Australia2005 24 11 2005 13 24 24 25 9 2005 24 11 2005 Copyright © 2005 Charlton; licensee BioMed Central Ltd.2005Charlton; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
There is little homogeneity of opinion in the chiropractic profession about its essence and identity. Matters compromising the establishment of a coherent identity include the issue of vertebral subluxation, philosophy, mercantilism, poverty of qualifications in some chiropractic college faculty, and lack of intellectual productivity in some chiropractic college faculty.
Discussion
The Chiropractic profession has mislabeled rhetoric, supposition and cant as philosophy, whilst showing sparse evidence for the existence of more than a few chiropractors writing in philosophy as a discipline. There is no evidence for "Chiropractic Philosophy".
I propose, however, that a better use of the discipline of philosophy can be of great use to the Chiropractic profession. Various thinkers throughout the ages have written about deduction, induction and falsificationism as methods to discover more reliably the nature of things in the world about us. Each method has strengths and frailties, but some of the latter are insurmountable for our purposes.
Summary
Using a contrivance of that method which seems most suited, sui generis, for the purpose, I propose a Chiropracticness Test as a tool to assist the search for essence and identity in Chiropractic.
==== Body
Background
More than 100 years after its foundation, the chiropractic profession still ruminates about its eternal internal debate on "what is chiropractic?"[1]. It seems to be a profession almost uniquely divided by mostly common purposes, and even others agree significant challenges lie ahead. [2]
Amongst the most important issues denying a thematically coherent identity to portray to ourselves, to our patients, to our students, to the academy and to the world at large are:
• The issue of vertebral subluxation, an interesting hypothesis for which there is almost no evidence, yet which much of the profession, including, mirabile dictu, its institutions and leaders, treat as reality (and some as catechismal chant); [3]
• Whilst there is real philosophical discourse in chiropractic publications [4-8] none of it yet is by those who teach it at chiropractic colleges. The obfuscation of the issue of philosophy as if there is an entity called "Chiropractic Philosophy", when there probably is not, is a major impediment to clarity of thinking in chiropractic; [9,10]
• The rise of mercantilism, for some, placing the patients' needs second to the commercial interests of the chiropractor; [11]
• The lack of disciplinary qualifications in, say, philosophy, for chiropractic college faculty who teach in subjects of that name;
• And the failure of so many chiropractic college faculty to publish suitably in the fields in which they teach, or at all. [12]
The foregoing, and more, denies the chiropractic profession the full cultural authority to be seen legitimately as the natural custodian of the milieu in which it thinks and operates. It has failed, as says Nelson et al [13], on the front of legitimacy. Our existing institutions "have not expressed a model of chiropractic that empowers the granting of cultural authority, sustained economic viability, and scientific integrity." Can the much abused "philosophy" as a discipline help us think more usefully about these shortcomings?
Discussion
Those immersed in the discipline of Philosophy as a vocation recognise several of its properties as characteristic: It deals with the clarification of important concepts or ideas and with clearer usage of key terms. It proceeds not by declaratory exposition, or by experimentation (there are no philosophy laboratories), but by rational endeavour, reasoning and argument. No matter how important an issue, or how wide its scope, a declaratory statement is not capable of being called philosophical unless it is defended or attacked by reasoning, not by recourse to authority, intuition or faith. Philosophy, indeed, is process.
Could Philosophy, real Philosophy, Philosophy as process, as a tool, help us here? Early last century, the great Cambridge philosopher Ludwig Wittgenstein called Philosophy a battle against the bewitchment of our intelligence by means of language [14], but I doubt he had the chiropractic profession in mind at the time.
Since this is mostly ultimately about the nature of the relationship between philosophy and science, it may help to think about how the scientific enterprise works. It seems likely that there re only two methods of rational justification: logic (deduction), and observation and experience (induction). Even these have problems. Deductive reasoning can provide no knowledge about the world around us. It offers but tautologies, or the implications of a given set of definitions and premises of, say, Euclidean Geometry. Even so, there is still no way to determine that the implications of any set of definitions and premises correspond to the complexity of observable reality.
The inductivist view is that one makes a suitably large number of observations and then distils a scientifically useful statement from the exercise. We say that if we collect large numbers of data under a wide variety of circumstances, we may endow our conclusions with greater confidence. Inductivism, the means by which most human biology research proceeds, has not escaped unscathed however. Herewith the thought processes of Bertrand Russell's inductivist turkey. It was, by all accounts, a very scientific bird, recording meticulously that it was fed every day at 9 am. It did not vary, whether the weather was bad or good, or by season. Eventually, very early one morning, the turkey felt able to claim "scientifically" "Because I have observed over a long time and under a wide variety of conditions, I now claim I will be fed every day at 9 am." Except it was Christmas Day, and at 9 am, the farmer killed the bird and the family ate it for lunch. No number of observations can exclude the possibility that a contrary future observation may render it invalid.
Because of this problem of induction exemplified by the turkey, some have felt that science needs a better way to proceed. Professor Sir Karl Popper, billed by his effusive biographer McGee as the greatest epistemologist since Aristotle, contrived falsificationism in response.
Popper made many contributions to understanding the nature of science, amongst them, the notion that no theory is ultimately provable. Perhaps his most significant contribution to the philosophy of science was his characterization of the scientific method. In The Logic of Scientific Discovery (1934; trans. 1959) [15], he criticized the prevailing view that science is fundamentally inductive in nature. Because he felt that reason operates primarily negatively, by criticism and refutation, he proposed a criterion of testability, or falsifiability, for scientific validity. Popper emphasized that a characteristic of a "scientific" theory is that it is vulnerable to refutation by observable events. If a hypothesis survives efforts to falsify it, it may be tentatively accepted.
Popper suggested we make precisely phrased risky conjectures about the world at large and make vigorous attempts to refute, or falsify, them. We can, he said, rely more on something we know to be false more than something we think is true because the latter may only have the status of something not yet proven wrong. If we have shown something to be false, we have learned something more useful. Truth lies somewhere in what is left. Thus, we aim for an asymptote of discovery, a verisimilitude.
Could, therefore, philosophy as process help us think about ourselves? What if we sought to phrase our claims for Chiropracticness and non-Chiropracticness in falsificationist-like prose as conjectures inviting refutation, and thus provide the meat and means for thinking about what is and is not Chiropractic? That we offer a pair of tests (The Chiropracticness Test) for what is and what is not Chiropractic at this time and in this place (it would vary).
When we wish to know whether any proposition is true, either of chiropractic or to our purpose as a profession, we must learn whether by conceivable variation of circumstances we can cause it to break down, either by its exclusion of what we think an essence of chiropractic, or its inclusion of what we are resolved to reject as inconsistent with that essence.
Summary
Such a test pair would be a new approach to philosophy in chiropractic, where rumination and rhetoric have held place so far, for we might now choose to use philosophy as a tool, better to understand ourselves. The Chiropracticness Test, then, uses falsificationist means to arrive at consensus, but in ways beyond the usual method of otherwise unstructured argument for or against a matter which catches our attention. Naturally, as with any science, future knowledge cannot be known now, or it would be present knowledge, not future, and future thinkers would need to apply the Test anew as times change, and the conclusions would certainly change in the light of new observations.
Our century-long failure to address philosophy in ways familiar to those who practice philosophy as a vocation is more an issue of abstinence than impotence: we can certainly do it.
Competing interests
The author(s) declare that they have no competing interests.
==== Refs
Nelson CF Lawrence DJ Triano JJ Bronfort G Perle SM Metz RD Hegetschweiler K LaBrot T Chiropractic as spine care: a model for the profession Chiropractic & Osteopathy 2005 13 9 16000175 10.1186/1746-1340-13-9
Cooper RA McKee HJ Chiropractic in the United States: Trends and Issues Milbank Q 2003 81 107 38 12669653 10.1111/1468-0009.00040
Keating JC Charlton KH Grod JP Perle SM Sikorski D Winterstein JF Subluxation: Dogma or Science? Chiropractic & Osteopathy 2005 13 17 16092955 10.1186/1746-1340-13-17
Donahue JD Metaphysics, rationality and science J Manipulative Physiol Ther 1994 17 54 5 8138734
Charlton KH Popper-Kuhn debate: a consideration of some of the implications for the philosophy of science and the chiropractic investigative community J Manipulative Physiol Ther 1988 11 224 7 3292689
Jamison JR Chiropractic holism: interactively becoming in a reductionist health care system Chiropr J Aust 1993 23 98 105
Coulter ID Alternative philosophical and investigative paradigms for chiropractic J Manipulative Physiol Ther 1993 6 419 25 8409791
Charlton KH Hit and Myth Chiropr J Aust 1991 21 58 61
Charlton KH Foolosofy The Australian Chiropractor 2004 magazine
Coulter ID Chiropractic A Philosophy for Alternative Health Care 1999 Oxford: Butterworth Heinemann
Charlton KH Silence is not golden: it's consent Chiropr J Aust 2003 33 Guest editorial 81 2
Wyatt LH Perle SM Murphy DR Hyde TE The necessary future of chiropractic education: a North American perspective Chiropractic & Osteopathy 2005 13 10 16001976 10.1186/1746-1340-13-10
Nelson op cit 2005
Wittgenstein L Pers DF,McGuiness BF Tractatus Logico-Philosophicus 1961 London: Routledge and Kegan Paul
Popper KR The Logic of Scientific Discovery 1959 New York: Harper and Rowe
|
16307687
|
PMC1308832
|
CC BY
|
2021-01-04 16:38:22
|
no
|
Chiropr Osteopat. 2005 Nov 24; 13:24
|
utf-8
|
Chiropr Osteopat
| 2,005 |
10.1186/1746-1340-13-24
|
oa_comm
|
==== Front
Clin Pract Epidemiol Ment HealthClinical Practice and Epidemiology in Mental Health : CP & EMH1745-0179BioMed Central 1745-0179-1-231628587910.1186/1745-0179-1-23ResearchA case control study on psychiatric disorders in Hashimoto disease and euthyroid goitre: not only depressive but also anxiety disorders are associated with thyroid autoimmunity Carta Mauro Giovanni [email protected] Maria Carolina [email protected] Bernardo [email protected] Andrea [email protected] Anna Rita [email protected] Fiora [email protected] Luca [email protected] Mariangela [email protected] Stefano [email protected] Division of Psychiatry, Department of Public Health, University of Cagliari, Italy2 Department of Internal Medicine, University of Cagliari, Italy2005 10 11 2005 1 23 23 2 10 2005 10 11 2005 Copyright ©2005 Carta et al; licensee BioMed Central Ltd.2005Carta et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Objective
To evaluate the association between mood and anxiety disorders in Hashimoto disease and Euthyroid Goitre in a case control study.
Methods
Cases included 19 subjects with Hashimoto disease in euthyroid phase, 19 subjects with euthyroid goitre, 2 control groups each of 76 subjects matched (4/1) according to age and sex drawn from the data base of a community based sample. Psychiatric diagnoses were formulated using the International Composite Diagnostic Interview Simplified, according to DSM-IV criteria. All subjects underwent a complete thyroid evaluation including physical examination, thyroid echography and measure of serum free T4 (FT4), free T3 (FT3), thyroid-stimulating hormone (TSH) and anti-thyroid peroxidase autoantibodies (anti-TPO). Results: Subjects with Hashimoto disease showed higher frequencies of lifetime Depressive Episode (OR = 6.6, C.L. 95% 1.2–25.7), Generalized Anxiety Disorders (OR = 4,9 Cl 95% 1.5–25.4) and Social Phobia (OR = 20.0, CL 95% 2.3–153.3) whilst no differences were found between subjects with goitre and controls.
Conclusion
The study seems to confirm that risk for depressive disorders in subjects with thyroiditis is independent of the thyroid function detected by routine tests and indicates that not only mood but also anxiety disorders may be associated with Hashimoto disease.
AutoimmunityAnxiety disordersGoitre Hashimoto diseaseMood disordersThyroid antibodies
==== Body
Introduction
Clinical and epidemiological studies seem to indicate that an association between high levels of thyroid autoantibodies may be implicated in the increased frequencies of mood disorders observed in thyroid disease, independently of impairment of thyroid function [1-3]. A study carried out by our group suggested a possible role of thyroid autoimmunity in the association between celiac disease and panic disorder and major depressive disorder [4].
The aim of the present study was to compare 2 clinical samples of subjects, one with Hashimoto disease (in euthyroid phase) and the second with euthyroid goitre versus controls drown from a community sample, in order to clarify whether the association of mood and anxiety disorder in Hashimoto disease is evident prior to impairment of thyroid dysfunction.
Methods
Design: case control study
Cases: 19 subjects, 18 females, mean age 39.7+/-12.6, with Hashimoto disease in euthyroid phase and 19 subjects, 16 females, mean age 38.1+/-10.4, with euthyroid goitre, all of whom were attending the Endocrinological Unit of the Department of Internal Medicine, University of Cagliari, Italy.
Two control groups, each of 76 subjects aged 18–64 years, were obtained by matching each "case" with four "controls" randomly selected on the basis of demographic characteristics (sex and age) from the data base of a larger population enrolled during a previous epidemiological study aimed at defining the prevalence of psychiatric [5] and thyroid diseases [6] in Sardinia. Controls were selected by a technique of randomisation after matching. For each case, a cell was formed containing all possible sex and age-matched controls (~1.0 year). The number of cells varied from 5 to 12 controls; a control was extracted at random from each cell. Exclusion criteria for control groups were the presence of abnormal values of serum free T4, free T3, thyroid-stimulating hormone (TSH) and anti-thyroid peroxidase autoantibodies (anti-TPO) as will be specified.
Two standardized forms were used to acquire information concerning: demographic data, state of health and use of social and health services. Psychiatric diagnosis was made using the Italian Simplified version of the Composite International Diagnostic Interview (CIDIS) [7]. The computer elaboration of data obtained enabled prevalence of psychiatric disorders according to DSM-IV [8] diagnostic criteria to be calculated.
All subjects underwent a complete thyroid evaluation including physical examination, thyroid echography and measure of serum free T4 (FT4), free T3 (FT3), thyroid-stimulating hormone (TSH) and anti-thyroid peroxidase autoantibodies (anti-TPO). FT4 and FT3 were measured by means of a chromatographic method based on separation of free T4 on Lisophase columns (Technogenetics, Milan, Italy; normal values: FT4 6.6–16 pg/ml; FT3 2.8–5.6 pg/ml); TSH was measured by a chemiluminescent method (Ortho-Clinical Diagnostics Amersham, U.K.) with normal values ranging from 0.3–3.0 μU/ml. Thyroid echography was performed using a "real time" echograph (ALOKA Mod SSD 500 with a small parts 7.5 Mhz sound. Anti-TPO, considered as the most sensitive and specific marker of thyroid autoimmunity [9] was determined by RIA (Sorin Biomedica Diagnostics, Saluggia, Italy) with a cut-off value of 20 IU/ml.
Results
Subjects with Hashimoto disease displayed high frequencies of lifetime Depressive Episodes (OR = 6.6, C.L. 95% 1.2–25.7), Generalized Anxiety Disorders (OR = 4,9 Cl 95% 1.5–25.4), Social Phobia (OR = 20.0, CL 95% 2.3–153.3) and Primary Sleep Disorders (OR = 20.0, CL 95% 2.3–153.3); a tendency towards an increased frequency of Panic Disorder was observed, although statistical significance was not reached (OR = 1.8, 0.1–24.6). No differences were found in the evaluation of lifetime prevalence of DSM-IV Psychiatric Disorders between patients with Euthyroid goitre and controls (Table 1).
Discussion
This study seems to indicate an association between the presence of a lifetime diagnosis of mood or anxiety disorder and Hashimoto disease without functional thyroid impairment. No difference was observed in mood and anxiety disorders between goitre cases and controls.
These findings are congruent with the results of a previous study [10] that indicated a risk for Depressive episode in subjects with antiTPO+ in a general population sample without selection by medical or psychiatric health facilities, and are consistent with several previous clinical studies providing evidence of a significant association of mood disorders or post-partum depression and symptomless autoimmune thyroiditis with or without sub-clinical hypothyroidism [11].
The above-cited community survey found that subjects with at least one lifetime diagnosis of anxiety disorders presented anti-TPO+ more frequently than subjects without mood or anxiety disorders. However, no specific anxiety diagnosis was found to be associated with anti-TPO+, although General Anxiety Disorder showed a strong trend toward association (P < 0.058). Anxiety Disorder Not Otherwise Specified was more frequently observed with anti-TPO+, but this is only a sub-threshold condition. A previous research carried out on celiac patients indicated thyroid autoimmunity as possible risk factor for panic disorder [3]. In the present study, the latter diagnosis was shown to be more frequent between Hashimoto disease and controls, although statistical significance was not reached.
Indeed, these 3 surveys seem to indicate a congruent trend of possible risk for anxiety disorders such Generalized Anxiety Disorder, Social Phobia and Panic Disorder in Hashimoto Disease.
To our knowledge, the data concerning the association between sleep disorders and Hashimoto disease is the first such evidence published in literature, although it may be consistent with a dysregulation of sleep architecture in subjects with sub-clinical thyroid impairment [12].
A sub-clinical dysfunction of axis thyrotropin releasing hormone (TRH) thyroid stimulating hormone (TSH) with consequent alteration of circadian rhythms of TSH has been hypothesized in several depressive disorders. Indeed, this hypothesis may explain why some forms of mood disorders were associated with anti-TPO+ without hypothyroidism, as defined by blood routine tests. A slight reduction in thyroid hormone secretion such as that found in sub-clinical hypothyroidism may affect cognition and mood [13]. At variance with other tissues which mainly rely on peripherally generated triiodothyronine, the brain utilizes preferentially circulating thyroxine directly secreted by the thyroid gland and may become hypothyroid before other organs [10].
Alternatively, autoimmunity may be implicated in some form of extra-thyroid disease associated (even indirectly) with the depressive symptomatology.
In Hashimoto disease the onset of vasculitis is frequently observed and would seem to be directly correlated to the duration of the illness [14].
In this disease brain perfusion modifications would appear to be of an aspecific nature; recently however, Spect examination revealed a marked compromising of the left cingulum posterioris, thus related in the etiology of associated affective disorders [14].
Recent evidence suggests that thyroid autoimmunity may be affected by the Hypothalamic-pituitary-adrenal axis (HPA) through the balance of proinflammatory and antiinflammatory cytokines [15,16]. In line with this view, the increased frequency of post-partum depression, associated to the fact that pregnancy would seem to be a "protected" period, could explain at least in part the consequences on thyroid autoimmunity elicited by HPA-related modifications to the immunitary axis. Indeed, similar phenomena are observed in both rheumatoid arthritis and multiple sclerosis and therefore Hashimoto vasculitis may also be involved [17].
Limitations
The potential of this study is limited by the small sample size, particularly regard to psychiatric diagnoses less frequently observed in the general population, such as Panic Disorder; the extension of the findings is therefore rather limited.
Conclusion
This study seems to indicate an association between the presence of a lifetime diagnosis of mood or anxiety disorder and sleep disturbance and anti-TPO+. The findings obtained suggest the need for further longitudinal studies aimed at clarifying the association between anxiety and depressive disorders and anti-TPO+.
==== Refs
Fountoulakis KN Iacovides A Grammaticos P Thyroid Function in Clinical Subtypes of Major Deoression BMC Psychiatry 2004 in press 15113438
Marangell B Callahan AM Mood disorders and the thyroid axis Current Opinion Psychiatry 1998 11 67 70 10.1097/00001504-199801000-00023
White AJ Barraclough B Thyroid disease and mental illness: a study of thyroid disease in psychiatric admissions J Psychosom Res 1988 32 99 106 10.1016/0022-3999(88)90093-1 3404495
Carta MG Hardoy MC Boi MF Mariotti S Carpiniello B Usai P Association between panic disorder, major depressive disorder and celiac disease. A possible role of thyroid autoimmunity Journal of Psychosomatic Research 2002 53 789 793 10.1016/S0022-3999(02)00328-8 12217453
Carta MG Kovess V Hardoy MC Morosini PL Murgia S Carpiniello B Psychiatric disorders in sardinian emigrants in Paris: a comparison with Parisians and Sardinians resident in Sardinia Social Psychiatry and Psychiatric Epidemiology 2002 37 112 117 10.1007/s001270200002
Loviselli A Oppo A Velluzzi F Independent expression of serological markers of thyroid autoimmunity and hepatitis C infection in general population J Endocrinol Invest 1999 22 660 665 10595828
Carta MG Carpiniello B Trudu MN Tarquini A Rudas N Aguglia E, Pascolo E La versione italiana della CIDI Simplified, uno studio di accuratezza e riproducibilità Metropoli e Oltre 1994 Trieste: Tentati
American Psychiatric Association Diagnostic and statistical manual of mental disorders, 4th edition (DSM-IV) 1994 Washington DC: American Psychiatric Association
Mariotti S Caturegli M Piccolo P Barbesino G Pinchera A Antithyroid peroxidase autoantibodies in thyroid diseases J Clin Endocrinol Metab 1990 71 661 669 2168432
Carta MG Loviselli A Hardoy MC Massa S Cadeddu M Sardu C Carpiniello B Dell'Osso L Mariotti S The link between thyroid autoimmunity (antithyroid peroxidase autoantibodies) with anxiety and mood disorders in the community: a field of interest for public health in the future BMC Psychiatry 2004 4 25 15317653 10.1186/1471-244X-4-25
Harris B Oretti L Lazarus J Parkes A John R Richards C Newcombe R Hall R Randomised trial of thyroxine to prevent postnatal depression in thyroid antibody positive Women Br J Psychiatry 2002 180 327 330 10.1192/bjp.180.4.327 11925355
Staner L Duval F Calvi-Gries F Morning and evening response to TRH and sleep EEG disturbances in major depressive disorders Prog Neuropsychopharmacol Biol Psychiatry 2001 25 535 47 10.1016/S0278-5846(00)00185-8 11370996
Haggerty JJ Prange AJ Borderline hypothyroidism and depression Annu Rev Med 1995 46 37 46 10.1146/annurev.med.46.1.37 7598471
Zetting G Asembaum S Feuger BJ Increased prevalence of subclinical brain perfusion abnormalities in patients with autoimmune thyroiditis: evidence of Hashimoto's encephalitis? Clinical Endocrinology 2003 59 637 644 10.1046/j.1365-2265.2003.01901.x 14616889
Elenkov IJ Chrousos GP Stress hormones, proinflammatory and anti-inflammatory cytokines, and autoimmunity Ann Y Acad Sci 2002 966 290 303
Tsigos Chrousos GP Hypothalamic-pituitary adrenal axis, neuroendocrine factors and stress J Psychosom Res 2002 53 865 71 10.1016/S0022-3999(02)00429-4 12377295
Elenkov IJ Wilder RL Bakalov VK Dimitrov MA Fisher S Crane M Kanik KS Chrousos GP IL-12, TNF-alpha, and hormonal changes during late pregnancy and early postpartum:implications for autoimmune disease activity during these times J Clin Endocrinol Metab 2001 86 4033 8 10.1210/jc.86.10.4933
|
16285879
|
PMC1308833
|
CC BY
|
2021-01-04 18:01:02
|
no
|
Clin Pract Epidemiol Ment Health. 2005 Nov 10; 1:23
|
utf-8
|
Clin Pract Epidemiol Ment Health
| 2,005 |
10.1186/1745-0179-1-23
|
oa_comm
|
==== Front
Clin Pract Epidemiol Ment HealthClinical Practice and Epidemiology in Mental Health : CP & EMH1745-0179BioMed Central 1745-0179-1-241628750610.1186/1745-0179-1-24Short reportShort report: autistic gastrointestinal and eating symptoms treated with secretin: a subtype of autism Pallanti Stefano [email protected] Stefano [email protected] Malfa Giampaolo [email protected] Marco [email protected] Rubbo Roberto [email protected] Giulia [email protected] Valentina [email protected] Department of Psychiatry, University of Florence, Italy2 Institute of Neuroscience, Florence, Italy3 SIRM (Italian Society for the study of Mental Retardation), Via Gordigiani, 58, 50127, Firenze, Italy2005 15 11 2005 1 24 24 12 9 2005 15 11 2005 Copyright ©2005 Pallanti et al; licensee BioMed Central Ltd.2005Pallanti et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Pervasive Developmental Disorders (PDD) are chronic, lifelong disorders for which there is as yet no effective cure, and medical management remains a challenge for clinicians. The current report describes two patients affected by autistic disorder with associated gastrointestinal symptoms.
They received multiple doses of intravenous secretin for a six-month period and were assessed with several specific outcome measures to evaluate drug effect.
The administration of secretin led to some significant and lasting improvement in only one case.
Gastroesophageal reflux may contribute to some of the behavioural problems and explain the effect of secretin since its suppressive effect on gastric secretion is well known. It is also true that autistic children with gastroesophageal reflux and a higher IQ could constitute a subtype which responds to secretin administration and that could be labelled as a "gastrointestinal subtype".
==== Body
Introduction
Pervasive Developmental Disorders (PDD) are chronic, lifelong disorders for which there is as yet no effective cure, and medical management remains a challenge for clinicians. In spite of improvements in some associated "problematic behaviors" with specific drugs, effective medical treatment for the core language- and social cognition-related symptoms are not available because the biology is not clearly understood and thus proper drug treatment has not been possible [1]. However, significant advances are being made towards understanding the mechanisms of the disorders, and major challenges lie ahead in evaluating the growing number of treatments for autism and in integrating the results of research into treatment and educational settings [2].
Since the experience of Horvath et al. [3] regarding secretin administration, with their report of "a dramatic improvement in the behavior of autistic children, manifested by improved eye contact, alertness, expansion of expressive language", and "relief of gastrointestinal symptoms", particular attention has been given to the potential role of this biological agent on autism.
Interest in secretin is also justified by many studies that have not only demonstrated the role of secretin as a classic hormone in the gastrointestinal system, but have firmly supported its neuropeptide role [4], since secretin has been shown to be capable of crossing the blood-brain barrier [5] and of depolarizing nucleus tractus solitarius neurons [6], activating brain regions including areas abnormal in autism [7].
Although preliminary reports [8] on secretin application initially generated enthusiasm, especially among parents of children with PDD [9], recent controlled studies seem to have dampened such enthusiasm (for a review, see [10]), although there is still some space for discussion.
A study by Sandler et al. [11] showed the lack of benefits in the treatment of core autistic symptoms with a single dose of secretin, though Sandler recognized that his study had several limitations: first, the follow-up period was short-term; second, only a single dose was administered; third, the diagnostic schedules used were not specific enough to measure the response to treatment.
Owley et al. [12] also conducted a double-blind versus placebo-controlled trial of porcine secretin for the treatment of autism, reaching the conclusion that there is no evidence of its efficacy.
Various studies conducted between 2000 and 2002 to assess the efficacy of a single dose of secretin on autistic features reported no significant effects [13]. All these studies adopted the DSM-IV criteria [14] and rating scales, which mainly focus on the so-called core autistic symptoms (only Corbett et al., [13] also used gastrointestinal measures to evaluate drug effects).
The wide range of autistic symptoms and their correlation with gastrointestinal functions, eating behavior and social interaction have not been focused on since Lightdale et al. [15] reported, following a single-blind, open-label pilot study, no effect of secretin in a five-week period on the language and behavior of 20 children with autistic and gastrointestinal symptoms.
Unis et al. [16] in a randomized double-blind, placebo-controlled study, reported no evidence that either biologic or synthetic secretin provided amelioration of symptoms beyond placebo. Likewise, Levy et al. [17] came to the conclusion that a single dose of secretin is not effective in changing behavior and communication in children with PDD if compared to placebo. In a controlled setting [18], parents of children with autism treated with a single dose of secretin were unable to distinguish the short behavioral effects of secretin from placebo.
These recent controlled studies seemed to spell the death knell for an unproven treatment that captured the public's imagination but found support in only a few positive reports [19].
However, there are several limitations in these very appreciable papers that leave some scope for us to report our observations. First, most of the studies were performed with a single infusion before the follow-up observational period. There are only three studies conducted on small samples [20] that report no evidence for the efficacy of repeated doses of secretin on the symptoms, language or cognitive functioning of children with autism. Second, the (paradoxically) large size of the sample when studying "categorized" autistic subjects leads to the risk of obtaining ungeneralizable data because the category of autism has a significant internal heterogeneity. Third, the wide range of typologies and the degrees of severity of symptoms have only been distinguished in with or without a widely defined condition labeled as "gastrointestinal symptoms", and conclusions cannot be drawn about hypothetically specific subtypes.
Methods
Given the controversial views surrounding the utility of this hormone in the treatment of autism, we decided to report our experience and considerations regarding two children with a diagnosis of autistic disorder according to DSM IV criteria. They were treated with secretin in order to assess any kind of clinical improvement; particular attention was devoted to the complexity of autistic eating behavior.
Participants
The first subject was a 9-year-old male, with a body weight of 35 kg, who was diagnosed as autistic at the age of two. We were able to confirm the previous diagnosis, as language and communication, social interaction and behavioral core symptoms were present; associated problems included a diet restriction and reflux esophagitis, and he cried at inappropriate times. He had an IQ of 99.
The second subject was a 7-year-old male with a body weight of 25 kg, who received his first diagnosis of autism at about the age of two. His IQ was 86. We confirmed the previous diagnosis, as he presented core symptoms and associated problems, especially gastrointestinal symptoms in the form of chronic diarrhea. The two boys were not taking other psychotropic medications during the study.
These two subjects were recruited consecutively, in a period when the data reported in literature about the potential effectiveness of secretin were promising. No other case was subsequently recruited (Table 1).
Table 1 Participant details.
Case Age (yr) Weight (kg) IQ SB-IS Gastrointestinal symptoms Core symptoms
1 9 35 99 Reflux esophagitis Strict diet Severe abnormal relationships Severe abnormal verbal communication Inappropriate crying Rocking Repetitive movements Abnormal adaptation to change Rituals Sleep problems
2 7 25 86 Chronic diarrhoea Severe abnormal relationships Severe abnormal verbal communication Routine Obsessive interests Repetitive movements Self-injurious behaviour Tantrums
Procedure
The method consisted of a single-blind protocol. The Stanford-Binet Intelligence Scale was employed before treatment in order to evaluate intelligence based on standardized criteria [21].
After obtaining informed consent from the parents, we administered aJapanese biological secretin, extracted from the duodenum of pigs, each vial containing 50 CU of secretin diluted in 2 ml. After a dose test of 1 CU over 1 minute, approximately 15 minutes prior to the full dose, carried out in order to determine possible allergic reactions, we administered the full intravenous secretin dosage of 2 CU/kg body-weight, given over a 1-minute period in a volume of 0.2 ml/kg, for each injection. The aim of the study was to administer 6 consecutive injections of secretin, one every 4 weeks. Each administration was carried out by an expert with proven experience of allergic anaphylactic reactions and resuscitation methods, because of the possibility that repeated use might result in an allergic response. Assessment included: 1. the Behavioral Summarized Evaluation [22], a 20-item scale that seemed a valid clinical tool to assess behavioral modifications and the evolution of the symptoms of children with autistic disorder; 2. the Clinical Global Impression Scale [23], a 3-item scale (Severity of Illness, Global Improvement, Efficacy Index) used to assess treatment response in psychiatric patients. We used these to measure eight separate features associated with autism (response to social interaction, social inition, use of speech, types of repetitive behavior, behavior problems, activity level, sleep problems, and digestive problems) on a standardized, seven-point Likert scale; 7 indicated "extremely ill", 4 "moderately ill" and 1 "normal" as regards Severity; 7 indicated "very much worse", 4 "no change", and 1 "very much improved" as regards Global Improvement; the Efficacy Index was measured on a four-point scale from "none" to "outweighs therapeutic effect"; 3. the Childhood Autism Rating Scale [24] which is the most widely used standardized instrument specifically designed to aid in the diagnosis of autism in young children [25], which includes 15 items (Relationships with People, Imitation, Affect, Use of Body, Relation to Non-human Objects, Adaptation to Environmental Change, Visual Responsiveness, Auditory Responsiveness, Near Receptor Responsiveness, Anxiety Reaction, Verbal Communication, Nonverbal Communication, Activity Level, Intellectual Functioning, and the Clinician's General Impression), with a symptom severity rating that makes it possible to use the scale for periodic monitoring and for assessing long-term outcomes.
An evaluator absolutely blind to the type of protocol assessed the children prior to and after the treatment, and in the follow-up with intervals of one day, one week, four weeks and six months after first treatment.
The pre-treatment assessment of the first boy yielded a BSE score of 75 (with a score of 4 on item 18, regarding eating disorders), a CGI score for severity of 6 and a CARS score of 53. Then he received 6 intravenous injections, one each month, each injection containing 75 CU of porcine secretin in 7 ml. He was evaluated with the three rating scales at intervals of one day, one week, and four weeks from the beginning of treatment. In this case it was possible to assess treatment response after six months from first injection and a month after last injection.
The second case had a BSE score of 63 (with a score of 4 on item 18, regarding eating disorders), a CGI score for severity of 6 and a CARS score of 52 in his pre-treatment evaluation. He then received 2 intravenous injections with an interval of a month, each injection containing 50 CU of porcine secretin in 5 ml; then he was evaluated with the three rating scales at intervals of one day, one week, and four weeks from the beginning of treatment. Treatment was suspended after the second injection since no effect was reported. The follow-up at six months is not therefore comparable but was conducted in any case since the second injection had been administered.
Results
In the first case, at interval 1 (after 1 day) there was no significant effect on the BSE, CGI (Severity was unchanged, the Global Improvement score was 4, i.e. no change, and the Efficacy Index score was 14, namely no treatment effect with no side effects), or CARS scores; at interval 2 (a week from first injection), we observed – clinically and through behavioral rating scales – a slight amelioration of his behavior, especially in alertness, expansion and efforts toward communication, while eye contact did not improve. In the same period parents reported an improvement in his diet, an associated problem; he began eating vegetables and pasta, which he had never wanted to eat before. He had a BSE score of 71, with a reduction from 4 to 3 on items 5, 6, 18 and 19. The CGI score for Severity did not change, the CGI score for Global Improvement dropped to 3, and the Efficacy Index score was 10 (mild effect with side effects that not interfere with the patient). His CARS score was 51.5 with a reduction of score at item 1, Relationships with People (4 to 3.5) at item 11, Verbal Communication (4 to 3.5), and at item 12, Nonverbal Communication (4 to 3.5). In the fourth week, at interval 3, these positive findings seemed to be still present; this was confirmed by a reduction of the BSE to 69, with a reduction on item 18 (eating disorders) from 3 to 1. We then proceeded with the second injection and since some slight amelioration of core symptoms was seen over the following four months we administered the other 4 planned intravenous injections. The symptoms follow-up after 6 months revealed enduring significant improvement with the same scores as at interval 3.
The second boy was evaluated prior to treatment and then at intervals of one day, one week, four weeks and six months after the treatment: no amelioration of core symptoms was noticed at any time of treatment in either eye contact, alertness or expansion of effort toward communication. Clinical observation confirmed the absence of positive results. We suspended treatment after the second injection in compliance with the parents' wishes.
At intervals 1 (one day from first injection), 2 (one week) and 3 (four weeks) there was no significant effect in the BSE, CGI (Severity was unchanged, the Global Improvement score was 4, i.e. no change and the Efficacy Index score was 14, i.e. no treatment effect with no side effects) or CARS (unvaried at 52) scores. Nor were any improvements reported after six months (Table 2).
Table 2 BSE score, CGI score for Severity, Global Improvement and Efficacy Index and CARS score at time 0, time 1 after one day, time 2 after one week, time 3 after four weeks and time 4 after six months.
Case BSE score CGI score S CARS score
0 1 2 3 4 0 1 2 3 4 0 1 2 3 4
1 75 71 69 69 69 6 6 6 6 6 53 53 51.5 51.5 51.5
GI 4 3 3 3
EI 14 10 10 10
2 63 63 63 63 64 6 6 6 6 6 52 52 52 52 52
GI 4 4 4 4
EI 14 14 14 14
Conclusion
The administration of secretin in our two young patients only led to some significant and enduring improvement on core symptoms in one case, where we observed significant changes in associated problems such as difficulties with toileting, sleeping and/or eating; laughing, crying or giggling at inappropriate times; response to touch, light, sound, taste or smells; unawareness of pain, heat or cold. In particular, we observed an evident amelioration of eating behavior associated with a stable global behavioral improvement.
This observation drew our attention to gastrointestinal symptoms and eating behavior, not only to dieting (even if in some cases it may be a relevant issue), in that the importance of social interaction and eating behavior has been pointed out by several studies [26]. Eating behavior in autistic children seems to be an intriguing subject for research, for several reasons:
a) because eating disorder, even when it is only considered as an associated problem, has important consequences on the health of these patients;
b) because of the important interaction between eating and social behavior; a recognized gastrointestinal disorder may contribute to the determination of behavioral problems in autistic patients.
Autistic disorder has always been considered to be substantially a neurobiological problem, but we cannot exclude different biological correlations, which could open up new perspectives in research and treatment of the disorder itself.
Besides, eating behavior is closely connected to attachment social behavior (i.e.: sucking is an act that includes the basis of eating and social behaviors, [27]). Studies conducted on animals [28] show that hormones such as oxytocin and vasopressin are involved in mediation at a central level of attachment behavior. A possible influence of secretin on autistic eating behavior could be mediated by the formation of unknown neuropeptides that, perhaps only in some children, finally have an influence on the construction of social behavior.
On the other hand Horvath et al. [29] found gastroesophageal reflux and reflux esophagitis to be the most frequently detected gastrointestinal abnormalities in children with autistic disorder (69.4%). In our report the child who improved with secretin treatment was the one characterized by reflux esophagitis, while the child who did not improve had chronic diarrhea but no gastric or esophageal reflux. It is known that secretin has a suppressive effect on gastric secretion [30]. Whether a low level of secretin may contribute to the high prevalence of acid reflux needs further investigations, but our report confirms the suggestion of Horvath & Perman [31] that gastrointestinal abnormalities may contribute to some of the behavioral problems, and the presence of esophagitis correlates well with the reported symptoms and may in part explain the sudden irritability, crying behavior and diet restriction as shown in our first case. In some way the behavioral problems connected to gastrointestinal abnormalities could represent a form of challenging behavior where a simple correction of dieting, or reflux reduction could determine great behavioral improvements [32]. It is also possible that autistic children with gastroesophageal reflux and esophagitis and higher IQ constitute a subtype, and probably respond better to secretin administration. Following this line of investigation, recent studies on autism suggest that there may be different subtypes of autism, with special reference to the "gastrointestinal subtype" [19].
We are justified in thinking that a more precise definition of each single case of autistic disorder would contribute to delineating possible subtypes of autistic disorder, which is currently a rather non-specific and over-inclusive diagnostic category. It is also true that many studies [33] show an increasing prevalence of PDD, especially among people with intellectual disability; this could in part be due to a better application of PDD DSM-IV criteria for autistic disorder or PDD not otherwise specified, and in part to increasing interest in the concept of the autistic spectrum. As a result of the relative non-specificity of the autistic disorder category, studies conducted on large samples may have the bias of studying non-homogenous subjects. Consequently, carefully described case reports could provide orientation regarding possible subtypes, so these can be described and recognized, and specific diagnostic criteria developed. The higher IQ index, and the presence or absence of some gastrointestinal symptoms in this case could represent indications of a somewhat specific subtype of autistic disorder, susceptible to specific treatments.
To conclude, secretin is not an effective cure for autism, as the extensive media attention suggested at the very beginning. However, observation of its efficacy on specific targets permits a number of reflections: it is only by using anecdotal reports and case observations showing reduction in the severity of specific symptom domains within autistic disorders, recognizing different specific subtype symptom domains, in order to elaborate good and valid study design with accurate selection of patients and specific outcome measures that it will be possible to obtain progress in the comprehension and cure of autism [34].
==== Refs
American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders 1994 4 Washington DC: APA 37 78 Return to citation in text: [14]
Banks WA Goulet M Rusche JR Niehoff ML Bois R Differential transport of a secretin analog across the blood-brain and blood-cerebrospinal fluid barriers of the mouse Journal of Pharmacology and Experimental Therapeutics 2002 302 1062 1069 Return to citation in text: [5] 10.1124/jpet.102.036129
Barthelemy C Adrien JL Tanguay P Garreau B Fermanian J Roux S Sauvage D Lelord G The Behavioral Summarized Evaluation: validity and reliability of a scale for the assessment of autistic behaviors Journal of Autism and Developmental Disorders 1990 20 189 204 Return to citation in text: [22]
Carey T Ratliff-Schaub B Funk J Weinle C Myers M Jenks J Double-blind placebo-controlled trial of secretin: effects on aberrant behavior in children with autism Journal of Autism and Developmental Disorders 2002 32 161 167 Return to citation in text: [13] 10.1023/A:1015493412224
Chey WY Chang TM Neural control of the release and action of secretin Journal of Physiology and Pharmacology 2003 Suppl 4 105 112 Return to citation in text: [4]
Chez MG Buchanan CP Bagan BT Hammer MS McCarthy KS Ovrutskaya I Nowinski CV Cohen ZS Secretin and autism: a two-part clinical investigation Journal of Autism and Developmental Disorders 2000 30 87 94 Return to citation in text: [13] 10.1023/A:1005443119324
Chung I Li P Lee K Chang T Chey WY Dual inhibitory mechanism of secretin action on acid secretion in totally isolated, vascularly perfused rat stomach Gastroenterology 1994 107 1751 1758 Return to citation in text: [30] 7958688
Coniglio SJ Lewis JD Lang C Burns TG Subhani-Siddique R Weintraub A Schub H Holden EW A randomized, double-blind, placebo-controlled trial of single dose intravenous secretin as treatment for children with autism Journal of Pediatrics 2001 138 649 655 Return to citation in text: [19] 11343038 10.1067/mpd.2001.112474
Coplan J Souders MC Mulberg AE Belchic JK Wray J Jawad AF Gallagher PR Mitchell R Gerdes M Levy SE Children with autistic spectrum disorders. II: parents are unable to distinguish secretin from placebo under double-blind conditions Archives of Disease in Childhood 2003 88 737 739 Return to citation in text: [18] 12876178 10.1136/adc.88.8.737
Corbett B Khan K Czapansky-Beilman D Brady N Dropik P Goldman DZ Delaney K Sharp H Mueller I Shapiro E Ziegler R A double blind, placebo-controlled crossover study investigating the effect of porcine secretin in children with autism Clinical Pediatry 2001 40 327 331 Return to citation in text: [13]
Dogrukol-Ak D Tore F Tuncel N Passage of VIP/PACAP/secretin family across the blood-brain barrier: therapeutic effects Current Pharmaceutical Design 2004 10 1325 1340 Return to citation in text: [5] 15134484 10.2174/1381612043384934
Dunn-Geier J Ho HH Auersperg E Doyle D Eaves L Matsuba C Orrbine E Pham B Whiting S Effect of secretin on children with autism: a randomized controlled trial Developmental Medicine and Child Neurology 2000 42 796 802 Return to citation in text: [13] 10.1017/S0012162200001481
Field D Garland M Williams K Correlates of specific childhood feeding problems Journal of Paediatrics and Child Health 2003 39 299 Return to citation in text: [32] 10.1046/j.1440-1754.2003.00151.x
Hollander E Phillips AT Yeh CC Targeted treatments for symptom domains in child and adolescent autism Lancet 2003 362 732 734 Return to citation in text: [34] 12957098 10.1016/S0140-6736(03)14236-5
Horvath K Papadimitriou JC Rabsztyn A Drachenberg C Tildon JT Gastrointestinal abnormalities in children with autistic disorder Journal of Pediatrics 1999 135 559 563 Return to citation in text: [29] 10547242 10.1016/S0022-3476(99)70052-1
Horvath K Stefanatos G Sokolski KN Watchtel R Nabors L Tildon JT Improved social and language skills after secretin administration in patients with autistic spectrum disorders Journal of the Association for Academic Minority Physicians 1998 9 9 15 Return to citation in text: [3] [8] 9585670
Horvath K Perman JA Autism and gastrointestinal symptoms Current Gastroenterology Reports 2002 4 251 258 Return to citation in text: [31] 12010627
Insel TR A Neurobiological Basis of Social Attachment American Journal of Psychiatry 1997 6 726 735 Return to citation in text: [27]
Insel TR Winslow JT Oxytocin and social attachment Giornale Italiano di Psicopatologia 1997 1 11 17 Return to citation in text: [28]
Kern JK Espinoza E Trivedi MH The effectiveness of secretin in the management of autism Expert Opinion on Pharmacotherapy 2004 5 379 387 Return to citation in text: [10] 14996634 10.1517/14656566.5.2.379
Kern JK Van Miller S Evans PA Trivedi MH Efficacy of porcine secretin in children with autism and pervasive developmental disorder Journal of Autism and Developmental Disorders 2002 32 153 160 Return to citation in text: [19] 10.1023/A:1015441428154
La Malfa GP Lassi S Bertelli M Salvini R Placidi GF Autism and intellectual disability: a study of prevalence on a sample of the Italian population Journal of Intellectual Disability Research 2004 48 262 268 Return to citation in text: [33] 15025669 10.1111/j.1365-2788.2003.00567.x
Lamson DW Plaza SM Transdermal secretin for autism – a case report Alternative Medicine Review 2001 6 311 313 Return to citation in text: [19] 11410075
Levy SE Repeated doses of porcine secretin did not improve symptoms, language or cognitive functioning in children with autism or autism spectrum disorder Evidence Based-Mental Health 2002 5 22 Return to citation in text: [20] 11915825
Levy SE Souders MC Wray J Jawad AF Gallagher PR Coplan J Belchic JK Gerdes M Mitchell R Mulberg AE Children with autistic spectrum disorders. I: comparison of placebo and single dose of human synthetic secretin Archives of Disease in Childhood 2003 88 731 736 Return to citation in text: [17] 12876177 10.1136/adc.88.8.731
Lightdale JR Hayer C Duer A Lind-White C Jenkins S Siegel B Elliott GR Heyman MB Effects of intravenous secretin on language and behavior of children with autism and gastrointestinal symptoms: a single-blinded, open-label pilot study Pediatrics 2001 108 90 Return to citation in text: [15] 11433059 10.1542/peds.108.5.e90
McQueen JM Heck AM Secretin for the treatment of autism The Annuals of Pharmacotherapy 2002 36 305 311 Return to citation in text: [10] 10.1345/aph.19113
Molloy CA Manning-Courtney P Swayne S Bean J Brown JM Murray DS Kinsman AM Brasington M Ulrich DD Lack of benefit of intravenous synthetic human secretin in the treatment of autism Journal of Autism and Developmental Disorders 2002 32 545 551 Return to citation in text: [13] 10.1023/A:1021202930206
National Institute of Mental Health CGI (Clinical Global Impression) Scale Psychopharmacological Bulletin 1985 21 839 843 Return to citation in text: [23]
Ng SS Yung WH Chow BK Secretin as a neuropeptide Molecular Neurobiology 2002 26 97 107 Return to citation in text: [4] 12392059 10.1385/MN:26:1:097
Owley T Steele E Corsello C Risi S McKaig K Lord C Leventhal BL Cook EH A double-blind, placebo-controlled trial of secretin for the treatment of autistic disorder Medscape General Medicine 1999 6 2 Return to citation in text: [12]
Owley T McMahon W Cook EH Laulhere T South M Mays LZ Shernoff ES Lainhart J Modhal CB Corsello C Ozonoff S Risi S Lord C Leventhal BL Filipeck PA Multisite, double-blind, placebo-controlled trial of porcine secretin in autism Journal of the American Academy of Child and Adolescent Psychiatry 2001 40 1293 1299 Return to citation in text: [13] 10.1097/00004583-200111000-00009
Palermo MT Curatolo P Pharmacologic treatment of autism Journal of Child Neurology 2004 19 155 164 Return to citation in text: [1] 15119475
Patel NC Yeh JY Shepherd MD Crismon ML Secretin treatment for autistic disorder: a critical analysis Pharmacotherapy 2002 22 905 914 Return to citation in text: [10] 12126223 10.1592/phco.22.11.905.33622
Perry R Bangaru BS Secretin in autism Journal of Child and Adolescent Psychopharmacology 1998 8 247 248 Return to citation in text: [8]
Roberts W Weaver L Brian J Bryson S Emelianova S Griffiths AM MacKinnon B Yim C Wolpin J Koren G Repeated doses of porcine secretin in the treatment of autism: a randomized placebo-controlled trial Pediatrics 2001 107 71 Return to citation in text: [20] 10.1542/peds.107.5.e71
Sandler D Sutton A DeWeese J Girardi A Sheppard V Bodfish J Lack of benefit of a single dose of synthetic human secretin in the treatment of autism and pervasive developmental disorder New England Journal of Medicine 1999 341 1801 1806 Return to citation in text: [11] 10588965 10.1056/NEJM199912093412404
Schopler E Reichler RJ DeVellis RF Daly K Toward objective classification of childhood autism: Childhood autism rating scale (CARS) Journal of Autism and Developmental Disorders 1980 10 91 103 Return to citation in text: [25]
Schopler E Reichler RJ DeVellis RF Daly K The Childhood Autism Rating Scale 1988 Los Angeles: Western Psychological Services Return to citation in text: [24]
Sponheim E Oftedal G Helverschou SB Multiple doses of secretin in the treatment of autism: a controlled study Acta Paediatrica 2002 91 540 545 Return to citation in text: [20] 12113323 10.1080/080352502753711669
Stokstad E Development. Desperate parents spark search for new treatment Science 2001 294 37 Return to citation in text: [9] 11588235 10.1126/science.294.5540.37
Thorndike RL Hagen EP Sattler JM The Stanford-Binet Intelligence Scale, Fourth Edition: Guide for administering and scoring 1986 Chicago: Riverside Return to citation in text: [21]
Unis AS Munson JA Rogers SJ Goldson E Osterling J Gabriels R Abbott RD Dawson G A randomized, double blind, placebo controlled trial of porcine versus synthetic secretin for reducing symptoms of autism Journal of the American Academy of Child and Adolescent Psychiatry 2002 41 1315 1321 Return to citation in text: [16] 10.1097/00004583-200211000-00012
Vitousek KB Ewald LS Segal Z, Blatt S Self-representation in eating disorders: a cognitive perspective The self in emotional disorders: cognitive and psychodynamic perspectives 1993 New York: Guilford Press 221 257 Return to citation in text: [26]
Volkmar FR Lord C Bailey A Schultz RT Klin A Autism and pervasive developmental disorders Journal of Child Psychology and Psychiatry 2004 45 135 170 Return to citation in text: [2] 10.1046/j.0021-9630.2003.00317.x
Welch MG Keune JD Welch-Horan TB Anwar N Anwar M Ruggiero DA Secretin activates visceral brain regions in the rat including areas abnormal in autism Cellular and Molecular Neurobiology 2003 23 817 837 Return to citation in text: [7] 10.1023/A:1025013322194
Yang B Goulet M Boismenu R Ferguson AV Secretin depolarizes nucleus tractus solitarius neurons through activation of a nonselective cationic conductance American Journal of Physiology – Regulatory, Integrative and Comparative Physiology 2004 286 927 934 Return to citation in text: [6] 10.1152/ajpregu.00600.2003
|
16287506
|
PMC1308834
|
CC BY
|
2021-01-04 18:01:02
|
no
|
Clin Pract Epidemiol Ment Health. 2005 Nov 15; 1:24
|
utf-8
|
Clin Pract Epidemiol Ment Health
| 2,005 |
10.1186/1745-0179-1-24
|
oa_comm
|
==== Front
Clin Pract Epidemiol Ment HealthClinical Practice and Epidemiology in Mental Health : CP & EMH1745-0179BioMed Central 1745-0179-1-251630573810.1186/1745-0179-1-25ResearchClinical and Social Outcomes five years after closing a mental hospital: a trial of cognitive behavioural interventions Mastroeni Antonino [email protected] Carla [email protected] Esterina [email protected] Francesco [email protected] Elena [email protected] Ian RH [email protected] Department of Mental Health, Appiano Gentile, Como, Itlay2 Department of Psychiatry, University of Auckland, New Zealand and ARIETE, 06050 Mercatello (PG), Italy2005 23 11 2005 1 25 25 4 10 2005 23 11 2005 Copyright ©2005 Mastroeni et al; licensee BioMed Central Ltd.2005Mastroeni et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.Background
To investigate the outcome of patients transferred from hospital to community care in Como, Italy after 6 months intensive psychosocial rehabilitation prior to discharge.
Method
All 149 residents with a primary psychiatric diagnosis were assigned to receive either a 6-month pre-discharge course of goal-oriented rehabilitation, (IT), or routine management, (RT). BPRS and GAF ratings were made by blind, independent assessors before and at 12, 24, 36, 48, and 60 months after discharge and the results examined with repeated measures analysis of variance.
Results
Overall change in residence was achieved without any major detriment to the health and welfare of most patients. The cohort of patients who received intensive rehabilitation, (IT), prior to discharge showed significantly lower impairment and disability throughout the five years compared to the cohort receiving routine management, (RT), prior to discharge. Total BPRS scores remained significant when initial differences in the cohorts were covaried, whereas GAF failed to remain significant (p = 0.051).
Conclusion
The treatment provided prior to transfer from long-stay hospital to community residence may have long-term clinical benefits for chronically disabled patients.
Mental Hospital ClosureEvidence-Based TreatmentControlled trial
==== Body
Introduction
Despite overwhelming evidence that institutional care is ineffective and often harmful for chronically impaired psychiatric patients, most policies to close long-term mental hospitals have been politically rather than professionally driven. Nowhere has that been more evident than in Italy. In 1978 the Italian Law 180 prevented the admission of any new cases to long-stay hospitals. This was followed in 1994 and 1995 by further national and local laws that aimed to accelerate the closure of mental hospitals that had been progressing very slowly [1]. The most recent laws fined Local Health Units and Hospitals if they did not close their mental hospitals and relocate patients to community housing before the end of 1999. One-third of the residents of these dilapidated hospitals were mentally ill, while another third were elderly or demented, and the remainder intellectually handicapped. Thus, the program to relocate patients to community housing was complex.
The benefits of community living for long-term severely disabled mentally disordered people have been fiercely debated [2-4]. A controversial problem has been the management of those behaviourally disturbed patients who are prone to violence or sexual misbehaviour who have been rehoused without the 24-hour supervision provided in hospitals [5]. Few prospective surveys have documented the process of closing mental hospitals and much of the debate has centred on media presentations associated with rare incidents of criminal violence or problems of homelessness. The most comprehensive study, the Friern Hospital Project, was carried out in London in the 1980s [2]. Extensive documentation provided a benchmark for patient relocation programmes of this kind. Other studies have provided less detailed reports of similar projects [6-8].
Although rehousing is the essential component of these hospital closure projects, the manner in which patients are prepared for this stressful lifestyle change should not be underestimated. Transitions of this kind offer an opportunity to review treatment and to ensure that evidence-based methods are applied for residual clinical and social morbidity, as well as to prevent exacerbations during the life change process and beyond.
We are aware of only one controlled trial of treatment methods used in the transition of long-term hospital patients to community care [9]. This study demonstrated that a structured educative programme using basic cognitive behavioural methods was more successful in achieving successful community tenure than milieu therapy or traditional supportive care.
The Como Project aimed to address both patient and professional competence for life and work in community settings. The key variable was training all staff in current evidence-based goal- and problem-oriented assessment, and biomedical and psychosocial treatments for all mental disorders. These methods provided the basis for preparing patients for community living, while preparing staff for major changes in their clinical practice. This paper outlines the process and provides a survey of the results achieved in a quasi-experimental study with 5-year follow-up.
Method
Overview (Figure 1)
Figure 1 On the 1st January 1997 415 patients were residents of Como Mental Hospital. 170 had a diagnosis of a primary mental disorder that was confirmed by a standardised review of current and past symptoms. In the two years prior to the hospital closure 73 of these cases had died, or were unable or unwilling to be interviewed. Of the remaining 97, 51 were assigned consecutively to an experimental rehabilitation program. It had been intended that all patients would complete this programme prior to discharge, but when the law dictated that the hospital close earlier than expected 46 cases had not yet begun the programme. No selection criteria were used to determine the order in which cases began the rehabilitation earlier. Thus, the two cohorts of patients could be compared in a naturalistic follow-up: 1) those receiving the hospital-based rehabilitation programme, and 2) those who were still awaiting the programme when the hospital closed. Patients in both groups were followed up every year after discharge to assess their clinical and social status.
It was hypothesised that this population suffered from disorders for which a programme of integrated evidence-based biomedical and psychosocial treatments would lead to stable reductions of clinical and social morbidity, with associated increased capacity to achieve personal goals such as independent living and the pursuit of satisfying occupational and social activities. To this end the professional staff began an intensive programme of training in evidence-based assessment and treatment strategies.
Staff characteristics
Key workers included nurses, social workers and occupational therapists, few of whom had previous specialized mental health training. They were supported by 7 psychiatrists and 1 clinical psychologist. As the process of discharging patients went on there was a parallel movement toward relocating staff to community settings.
Rehabilitation programme, IT
In 1997 a training project was established for all staff who were caring for mentally disordered residents. This training was part of the international Optimal Treatment Project [10]. The training was adapted to the needs of long-term disabled patients and included workshops on comprehensive standardized biomedical and psychosocial assessments, clarifying patients' personal goals, educating patients about their mental disorders and treatments, optimal pharmacotherapy, early warning signs of exacerbations, assertive community treatment and crisis management, enhancing interpersonal communication and social skills, enhancing personal self-care, structured problem solving and other cognitive behavioural strategies to aid coping with targeted residual psychotic, negative, anxiety and mood symptoms, as well as problems of substance abuse, anger and frustration. This intervention was termed Integrated Treatment, IT [see [10]].
The Italian versions of OTP professional manuals and patient guidebooks were the basis for the professional and patient training sessions [11]. These were educational lesson guides designed to be easily followed by professionals and patients alike. They were based on the principles of error-free learning and practical skills training targeted to the explicit personal goals and key problems of each patient. After practice and discussion in individual and group sessions patients applied the strategies in their actual life situations and reported their outcomes at the next training session, where they received praise and encouragement for their efforts and further coaching to help achieve their goals to the level that they considered satisfactory, before moving on to another goal that they considered likely to improve their current life quality. A total of 100 hours of workshop training and supervision was provided over two-years.
They lived together in groups of 8–10, either in "apartments" or in other residential facilities within confines of the hospital, without restrictions, except for the need to follow straightforward cohabitation rules that were agreed among fellow residents. They were able to practice their skills and work on their goals together with daily staff coaching. Treatment was completed at discharge from the hospital. Although efforts were made to continue this treatment once patients were resident in the community, often this was not feasible, either because staff in the community residences and mental health services were not trained in the methods, or more commonly because managers favoured other approaches.
Routine Treatment, RT
Patients awaiting IT were treated by the same group of professionals. Pharmacotherapy, nursing care and occupational therapy was provided within a supportive problem-oriented framework. However, no structured psychosocial assessment or treatment protocols were provided.
Assessment
All patients were assessed by an independent assessor, who was blind to treatment allocation, at the start of the project and at yearly intervals thereafter. The assessor was trained to administer the following ratings to an intra-class reliability of at least 0.80.
Clinical
The BPRS-24 [12] was used to assess the severity of psychiatric symptoms.
Psychosocial Functioning
The Italian version of Global Assessment of Functioning, GAF, [13] was used to assess psychosocial functioning.
Data Analysis
In addition to descriptive data, repeated measures analysis of variance using SPSS-PC 10.1.0 was conducted to evaluate trends over time on the entire cohort and interactions between the two treatment groups. In order to compensate for any non-random differences between the two groups the duration of illness, age, gender as well as initial ratings on each specific rating scale were included as covariates. An alpha of .05 was used to define statistical significance.
Results
Sample of residents who entered the project (Table 1)
Table 1 Characteristics of IT and Routine Treatment patients at the beginning of the hospital closure project
IT group RT group
N = 51 46
Age (mean years and range) 57 (35–70) 58 (40–73)
Gender:
Male (%) 21 (41%) 33 (72%)
Female(%) 30 (59%) 13 (18%)
Duration of Illness (mean years and range) 23 (6–41) 27 (8–40)
Duration of this hospital admission (mean years and range) 16 (6–29) 19 (7–40)
Diagnosis: DSM-IV:
Schizophrenic Disorders (%) 38 (74%) 37 (81%)
Affective Disorders (%) 7 (14%) 2 (4%)
Anxiety and Personality Disorders (%) 1 (2%) 2 (4%)
Substance Abuse (%) 1 (2%) 5 (11%)
Mental retardation (%) 4 (8%) -
Died during follow-up period 8 (16%) 6 (13%)
5th year evaluation not available 1 (stroke) 1 (missed interview)
Complete assessments throughout 5 years 42 (82%) 39 (85%)
The characteristics of the sample are documented in Table 1. The mean age was 58 years and was similar in both groups. There was an excess of men in the overall sample (54 vs 43). The proportion of men to women was significantly greater in the routine treatment, RT than in IT. 14 patients died after discharge (8 in IT and 6 in RT). All deaths were from natural causes. Two other cases were unable to complete the 5 annual assessments; 1 IT case had a stroke, and 1 RT case was admitted to forensic hospital in another area.
More than two-thirds had a DSM-IV diagnosis of a schizophrenic disorder. These were equally distributed between the treatment groups. The mean duration of mental disorders was 25 years (range 6 to 41 years). Although the duration of cases entering IT was significantly lower, the respective mean durations of 23 and 27 years would appear to have limited clinical significance. More than 80% had been mentally ill for more than 20 years, and half of these had been resident in the hospital continuously for the past 20 years.
Thus, apart from the disparity of the gender mix, the two cohorts appeared closely matched.
Residence
Four patients (4.1%) returned to their own homes or to live with family or friends; 89 (92%) went to live in 13 specialized residences for psychiatric patients. Almost all patients were able to choose their own place of residence and most were happy with the choice. However, probably owing to the precipitous hospital closure, there were many subsequent changes, with one quarter changing residence at least once in the 5 years after discharge.
Homelessness
One RT patient with a history of vagrancy became homeless after repeatedly refusing housing. According to his wishes he was given social care and shelter, including admissions to the general hospital psychiatric ward when he requested it. He continued to receive regular treatment most of the time, but missed most assessment interviews.
Criminality and Behavioural Problems in the Community
The same homeless patient was arrested twice. Once for threatening behaviour and carrying a weapon (a knife), for which he was not charged. One the second occasion he was charged with burglary and admitted to a forensic hospital. No other problems were reported to police, community services or local authorities.
Clinical Outcome in the quasi-experimental study (Table 2)
Table 2 Clinical and Social Outcome of 81 cases who completed all six baseline and follow-up assessments
BPRS total GAF
Assessment time (months) Treatment group N Mean Std. Deviation Mean Std. Deviation
0 otp 42 37.02 11.46 45.14 10.92
rt 39 42.05 17.25 37.85 11.68
total 81 39.44 14.66 41.63 11.80
12 otp 42 36.43 14.15 50.69 13.83
rt 39 42.44 18.29 41.26 13.71
total 81 39.32 16.45 46.15 14.48
24 otp 42 40.17 13.72 49.24 13.24
rt 39 48.00 20.27 39.69 13.70
total 81 43.94 17.52 44.64 14.22
36 otp 42 40.02 12.00 52.26 12.64
rt 39 48.64 19.18 41.54 12.61
total 81 44.17 16.35 47.10 13.66
48 otp 42 35.79 11.15 53.40 12.51
rt 39 45.59 18.41 40.72 13.10
total 81 40.51 15.78 47.30 14.23
60 otp 42 37.62 12.77 53.90 14.63
rt 39 46.33 18.54 40.77 13.69
total 81 41.81 16.31 47.58 15.57
Eighty-one patients (82%) completed all 6 BPRS interviews; 42 in the IT group and 39 in RT.
On the sum of BPRS items the total cohort of discharged patients showed significant deterioration over the 5-year follow-up period (repeated measures ANOVA with Greenhouse-Geisser correction for lack of sphericity: F = 4.22; df 3.34, 271.74, p = .004). The IT group improved after the program in hospital, then gradually deteriorated in the second and third years in the community before regaining their baseline level. By contrast, the RT group showed significant deterioration from baseline during years 2 and 3 and improved somewhat during years 4 and 5, but did not regain their baseline level.
Repeated measures analysis of variance showed a significant group × time interaction (F = 6.764; df 1,79; p = .011). This significant interaction remained when baseline BPRS, age, sex and duration of illness were entered as covariates (F = 4.802; df 1,75; p = .032). This supports the observation that the IT group remained more stable over the follow-up period whereas a deteriorating trend was observed for RT.
Eighty-two cases completed all GAF assessments. The entire cohort showed a modest but significant trend to improve over the follow-up period (repeated measures ANOVA with Greenhouse-Geisser correction for lack of sphericity: F = 10.04; df 3.622, 293.36; p = .001). Both groups followed this trend and showed significant improvements with time. Most improvement occurred during the first year after discharge. IT cases improved most during the training programme and showed smaller improvements after discharge. RT cases were not assessed at discharge, but showed significant improvements in functioning during the first year of community living, but these were less well sustained.
Repeated measures analysis of variance showed a significant group × time interaction (F = 15.99; df 1,79; p < .001), but on this occasion it just failed to remain significant when the baseline GAF assessment, age, sex and duration of illness were all entered as covariates (F = 3.95; df 1,75; p = .051).
In order to clarify the clinical significance of these findings we constructed an "index of recovery". Cases considered to have made a good recovery were those who had a BPRS total score of between 24 (the minimum) and 30; and in addition had a GAF score of at least 60. At the baseline assessment 8.5% of cases allocated to the IT interventions and 4.8% of those receiving RT were had achieved a good recovery. At 5 years 33.3% of the IT and 10.3% of RT cases met the good recovery threshold. This advantage for IT was significant (Fisher's exact test: p = .016, two-sided).
In order to confirm that the main factor associated with the better clinical and social outcome for the IT cohort was the programme of evidence-based interventions received before discharge, we conducted an ordinal multiple regression analysis with the index of recovery as the dependent variable and the following variables entered into the equation: treatment group, diagnosis (schizophrenia vs. other), age (older or younger than the median age of 62 years), gender, and duration of illness (greater or less than 29 years median). The only variable that was statistically significant in the regression equation was the treatment group allocation. This added support to the hypothesis that the pre-discharge period of intensive rehabilitation contributed to the benefits of those allocated to IT.
Discussion
The Como Project enabled a large mental hospital in North Italy that provided care for equal proportions of neurologically and psychiatrically disabled patients to be closed over a 3-year period with minimal difficulty. A five-year follow-up showed that most psychiatric patients were resettled in residences in the community that appeared to provide similar social and medical support to that which they had received in the long-stay wards. This could be considered a success, as there was little evidence of serious problems either during or after resettlement, including a low rate of criminal or antisocial activity, death and suicide rates that remained stable and lower than those found in the closure of the mental hospitals in London [14].
While providing adequate humane nursing home facilities for elderly demented patients, and adults with developmental disorders may be considered a success, the same results cannot be hailed as a major achievement for rehabilitation. In the past few decades substantial progress has been made in the biomedical and psychosocial treatment of mental disorders, with associated increased rates of recovery, even for cases who do not receive such treatment until relatively late in the course of their disorders [15]. Although the entire cohort of cases with psychiatric disorders did not show any notable clinical or social deterioration over the 5-year period of assessment at best the trend was for clinical and social stability rather than recovery. However, for the cases that received integrated evidence-based treatment for 6 months prior to discharge, statistical and clinically significant reductions in morbidity were evident. Some gains were lost when this treatment did not continue after hospital discharge. However, after 5 years the benefits of this relatively brief intervention were still clearly evident, and one-third had achieved a good recovery from clinical and social morbidity. In contrast only 10% of those in the comparison group had achieved such a status. These latter results are consistent with the findings of other hospital closure programmes that had not introduced specific evidence-based rehabilitation strategies [16,17].
It is clear that recovery from the symptoms and associated disability of mental illness is a slow process that demands continuous optimal treatment for many years. Residential alternatives to long-stay hospital wards may prove less expensive and reduce the alienation of the severely mentally from community resources and opportunities. But unless they are associated with an improvement in the quality of treatment that is provided, many will remain mere asylums in the community that may lead to increased stigma for such disabled people and calls to re-open the large institutions [18].
Extreme care must be taken in the interpretation of these results, because this naturalistic project had many limitations. Although there was no overt bias in the selection of cases for the intensive rehabilitation programme, the sampling was convenience-based and not random. Matching was good, but not perfect, with traditional prognostic factors favouring the IT cohort. Statistical corrections of baseline differences between the two groups did not change the results substantially. However, replications with more rigorous methodology are essential before it can be concluded that the psychosocial interventions used in this project are efficacious in this group of long-term patients both in facilitating relocation to community care, and in enhancing clinical and social recovery. One such project that uses an identical rehabilitation approach is in progress in Koriyama, Japan, with preliminary results suggesting similar benefits [19]. To date hospital closure programmes have been considered successful if patients have managed to merely relocate to community housing without excessive clinical exacerbations, excessive readmissions to acute or long-term hospital facilities, or involvement with the criminal justice system [2,5,17]. This report is one of the few that has reported positive outcomes from such a project. It is evident that mere re-housing achieves limited benefits, and may even be associated with some deterioration in many cases. However, when this major upheaval in the lives of vulnerable people is accompanied by an effort to provide state-of-the-art biomedical and psychosocial treatment such programmes may contribute to significant long-term clinical and social benefits [9,20,21].
An additional major positive effect has been the relative ease of transition to work in community services of the professional staff that participated in the project. The treatment strategies they learned and applied with limited success to the complicated cases in the hospital were equally useful for cases attending community-based services. The greater success encountered in these settings created considerable enthusiasm for the new work environment. All too often efforts to establish effective community-based services prove difficult when staff lack competence in evidence-based treatment strategies for patients they are expected to treat. The Como Project appeared to circumvent this problem through the training and experiences provided during the hospital closure process. In the same manner that the patients need long-term efficacious treatment, staff need continued supervision and upgrading of their therapeutic competence [22]. This is a problem that is now been tackled in the community-based services that have replaced the Como Mental Hospital.
Authors' contributions
AM designed the project, administered all procedures, analysed and intepreted the data and prepared the manuscript.
CB administered the project throughout, managed the database, analysed the data and assisted in preparation of the manuscript.
EP organised the staff training, administered the clinical procedures, assisted in the preparation of the manuscript.
FG assisted in the staff training, administered the clincial procedures, translated work materials and guidebooks
EL prepared the database, assisted in the data analysis
IRHF conducted the training and supervision, analysed the data, assisted in preparing the manuscript
Acknowledgements
Funding for this project was provided by the Regione Lombardia.
==== Refs
de Girolamo G Cozza M The Italian psychiatric reform. a 20-year perspective Int J Law Psychiatry 2000 23 197 214 10.1016/S0160-2527(00)00030-3 10981267
Leff J Care in the community. Illusion or reality? 1997 Chichester, John Wiley & Sons
Lamb R Lessons learned from deinstitutionalization in the US Br J Psychiatry 1993 162 587 592 8149108
Mollica RF From asylum to community. The threatened disintegration of public psychiatry New Eng J Med 1983 308 367 373 6823241
Gudeman JE Shore MF Beyond deinstitutionalization. A new class of facilities for the mentally ill New Eng J Med 1984 311 832 836 6472385
Braun P Kochansky G Shapiro R Greenberg S Gudeman JE Johnson S Shore MF Overview: Deinstitutionalization of psychiatric patients. A critical review of outcome studies Am J Psychiatry 1981 138 736 749 7018273
Hobbs C Tennant C Rosen A Newton L Lapsley HM Tribe K Brown JE Deinstitutionalization for long-term mental illness: a 6-year evaluation Aust N Z J Psychiatry 2002 36 60 66 10.1046/j.1440-1614.2002.00984.x 11929439
Kaiser W Hoffmann K Isermann M Priebe S Long-term patients in supported housing after deinstitutionalisation – part V of the Berlin Deinstitutionalisation Study Psychiatric Praxis 2001 28 235 243 10.1055/s-2001-15577
Paul GC Lentz RJ Psychosocial Treatment of Chronic Mental Patients 1977 Cambridge MA: Harvard University Press
Falloon IRH OTP Collaborators Optimal treatment for psychosis in an international multisite demonstration project Psychiatr Serv 1999 50 615 618 10332895
Falloon IRH OTP Collaborators Trattamento Integrato per la Salute Mentale 2000 Salerno: Ecomind Publications
Ventura J Green MF Shaner A Liberman RP Training and quality assurance with the Brief Psychiatric Rating Scale: 'The drift busters" Int J Methods Psychiatr Res 1993 3 221 244
American Psychiatric Association Diagnostic and Statistical Manual of Mental Disorders 1993 4 APA: Washington DC
Trieman N Leff J Glover G Outcome of long stay psychiatric patients resettled in the community: prospective cohort study BMJ 1999 319 13 16 10390451
Falloon IRH Montero I Sungur M Mastroeni A Malm U Implementation of evidence-based treatment for schizophrenic disorders: two-year outcome of an international field trial of optimal treatment World Psychiatry 2004 3 104 109
Leff J Trieman N Long stay patients discharged from psychiatric hospitals Br J Psychiatry 2000 176 217 223 10.1192/bjp.176.3.217 10755067
Barbato A D'Avanzo B Rocca G Amatulli A Lampugnani D A study of long-stay patients resettled in the community after closure of a psychiatric hospital in Italy Psychiatr Serv 2004 55 67 70 10.1176/appi.ps.55.1.67 14699203
de Girolamo G Picardi A Micciolo R Falloon I Fioritti A Morosini P PROGRES Group Residential care in Italy. National survey of non-hospital facilities Br J Psychiatry 2002 181 220 225 10.1192/bjp.181.3.220 12204926
Mizuno M Sakuma K Ryu Y Takebayashi T Murakami M Falloon IRH Kashima H The Sasagawa Project: rationale, methods, and recruitment Keio Med J 2005 54 95 101
Thornicroft G Bebbington P Deinstitutionalisation – from hospital closure to service development Br J Psychiatry 1989 155 739 753 2695205
Economou M Palli A Falloon IRH Violence, misconduct and schizophrenia: Outcome after 4 years of optimal treatment Clinical Practice and Epidemiology in Mental Health 2005 1 3 10.1186/1745-0179-1-3
Senn V Kendal R Trieman N The TAPS project 38: level of training and its availability to carers within group homes in a London district Soc Psychiatry Psychiatr Epidemiol 1997 32 317 322 10.1007/BF00805435 9299924
|
16305738
|
PMC1308835
|
CC BY
|
2021-01-04 18:01:02
|
no
|
Clin Pract Epidemiol Ment Health. 2005 Nov 23; 1:25
|
utf-8
|
Clin Pract Epidemiol Ment Health
| 2,005 |
10.1186/1745-0179-1-25
|
oa_comm
|
==== Front
Cost Eff Resour AllocCost effectiveness and resource allocation : C/E1478-7547BioMed Central London 1478-7547-3-111628864610.1186/1478-7547-3-11ResearchThe costs of reducing loss to follow-up in South African cervical cancer screening Goldhaber-Fiebert Jeremy D [email protected] Lynette E [email protected] Souza Michelle [email protected] Thomas C [email protected] Louise [email protected] Sue J [email protected] Harvard Initiative for Global Health, Harvard University, Massachusetts, USA2 Department of Obstetrics and Gynecology, University of Cape Town, South Africa3 Department of Pathology, College of Physicians and Surgeons of Columbia University, New York, USA4 Gertrude H. Sergievsky Center, College of Physicians and Surgeons, and Division of Epidemiology, Joseph L. Mailman School of Public Health, Columbia University, New York, USA2005 15 11 2005 3 11 11 6 7 2005 15 11 2005 Copyright © 2005 Goldhaber-Fiebert et al; licensee BioMed Central Ltd.2005Goldhaber-Fiebert et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
This study was designed to quantify the resources used in reestablishing contact with women who missed their scheduled cervical cancer screening visits and to assess the success of this effort in reducing loss to follow-up in a developing country setting.
Methods
Women were enrolled in this Cape Town, South Africa-based screening study between 2000 and 2003, and all had scheduled follow-up visits in 2003. Community health worker (CHW) time, vehicle use, maintenance, and depreciation were estimated from weekly logs and cost accounting systems. The percentage of women who attended their scheduled visit, those who attended after CHW contact(s), and those who never returned despite attempted contact(s) were determined. The number of CHW visits per woman was also estimated.
Results
3,711 visits were scheduled in 2003. Of these, 2,321 (62.5%) occurred without CHW contact, 918 (24.8%) occurred after contact(s), and 472 (12.7%) did not occur despite contact(s). Loss to follow-up was reduced from 21% to 6%, 39% to 10%, and 50% to 24% for 6, 12, and 24-month visits. CHWs attempted 3,200 contacts in 530 trips. On average, 3 CHWs attempted to contact 6 participants over each 111 minute trip. The per-person cost (2003 Rand) for these activities was 12.75, 24.92, and 40.50 for 6, 12, and 24-month visits.
Conclusion
CHW contact with women who missed scheduled visits increased their return rate. Cost-effectiveness analyses aimed at policy decisions about cervical cancer screening in developing countries should incorporate these findings.
==== Body
Background
The vast majority of cervical cancer deaths occur among women in developing countries where screening has been largely unavailable [1]. The effectiveness of cervical cancer screening has been demonstrated by the dramatic decline of cervical cancer in developed countries in which programs relying on repeated cervical Pap smears have been successfully implemented [2,3].
Traditionally, conventional cervical cancer screening using cervical cytology requires up to three visits (screening, colposcopy/biopsy, and treatment). In developed country settings, HPV DNA testing has been proposed both as a primary screening test in older women and in conjunction with cervical cytology as triage for equivocal cytologic results [4,5]. Recently, others have proposed using either cervical cytology or HPV DNA testing in two-visit strategies that eliminate confirmation with colposcopy/biopsy prior to treatment, or using one-visit, "see and treat" strategies with visual inspection with acetic acid (VIA) [6-8]. Regardless of the screening test chosen, an important motivation for these alternative strategies is to reduce the screening program's susceptibility to loss to follow-up by reducing the number of visits at which loss to follow-up can occur [9-11].
A number of cost-effectiveness analyses (CEAs) have been conducted, comparing many of these screening strategies in developing country contexts [12-14]. One key finding is that for screening strategies that require women to return to the clinic multiple times, loss to follow-up drives down the effectiveness of screening, regardless of screening modality. Furthermore, the cost of screening accrues as each woman is screened. The primary benefit from screening, the prevention of cervical cancer, is only realized for those women with positive test results and true precancerous lesions who are ultimately treated.
In low resource settings, levels of loss to follow-up in cervical cancer screening programs where follow-up visits were scheduled four weeks or more after the initial visit range from 10% to 70% [10,15-21]. Efforts to reduce loss to follow-up and to maintain it at an acceptably low level are thus a key part of cervical cancer screening programs. Such efforts can be time-intensive and costly and do not guarantee that all women return to the clinic for follow-up. Quantifying the cost and effectiveness associated with achieving acceptably low levels of loss to follow-up is essential for providing an estimate of the investment necessary to achieve cervical cancer screening coverage at the population level as well as for estimating the true cost-effectiveness of cervical cancer screening in these settings.
We used year 2003 data from the Khayelitsha Cervical Cancer Screening Program (KCCSP), a multi-year, South African program, to examine the success of CHW home visits in reducing loss to follow-up as well as the extent and type of resources used in this effort.
Methods
Study setting
The Khayelitsha Cervical Cancer Screening Program encompasses a multi-year, multi-site study designed to evaluate the effectiveness of a variety of cervical cancer screening and treatment strategies among a largely poor, black, peri-urban, South African population in a real-world setting. Approximately 25% of participants live in informal settlements without basic services such as electricity and water, and another 45% live in informal settlements with some basic services. The remaining 30% live in formal settlements.
At each scheduled appointment, women are tested with cervical cytology, HPV DNA testing using Hybrid Capture II, and visual inspection with acetic acid. The program has generated data that support the effectiveness and cost-effectiveness of various screening strategies [11,22,23]. In an effort to minimize loss to follow-up, CHWs drive throughout the community to visit women in their homes if they have missed appointments. Home visits by CHWs are used because most participants do not have telephones.
Study population
All study participants who had appointments in 2003 were considered eligible for inclusion in the main analysis. Study participants had one or more of four different types of appointments scheduled for during this period: 6-month, 12-month, 24-month, and 36-month visits. We restricted our analysis of CHW home visit effectiveness and examined the first three types of appointments because the number of 36-month appointments in 2003 was small (n = 277) relative to 6-month (n = 919), 12-month (n = 1820), and 24-month (n = 980) appointments.
KCCSP was approved by the Institutional Review Boards of Columbia University and the University of Cape Town, and all study participants gave written informed consent.
Effectiveness
To assess the effectiveness of CHW home visits, we estimated the number of appointments that would have been missed due to loss to follow-up and the percentage of these appointments rescheduled and attended due to home visits. We accounted for the possibility that each participant might have more than one appointment within the period of this study and that CHWs might make more than one home visit per participant.
Participant study number, entries for all scheduled appointments during 2003, and the date(s) on which the participant arrived for these appointments were extracted from the main study database. This data was linked to data transcribed from weekly forms used by CHWs to record participant study numbers and the date of the CHW visits. Using the date of the scheduled appointment, the date of the CHW visit(s), and the date on which the participant actually visited the clinic, the appointment type for which the participant was visited was identified. Using this method, those participants who had been visited by CHW but never returned for their appointments were also identified. Similar estimates subcategorized by appointment type were also generated.
Costs
To estimate costs, we first identified the different types of resources used in the CHW home visits. Resources included the time the CHWs spent driving to and from as well as visiting participants who had missed appointments, fuel used during the trips to visit participants, and the maintenance and depreciation in the value of the vehicles. Next, the average monetary value for each resource type was estimated. Finally, the estimated quantity of each type of resource was multiplied by its estimated monetary value, and the results were then summed to calculate the total cost of the effort to reduce loss to follow-up.
The quantity of CHW time used for visits was derived from weekly reporting sheets that identified the amount of time spent making home visits each day, the study number of each participant visited, and the CHWs who went on each visit. The value of CHW time was estimated in two ways. First, the salary scale employed within the study was used to estimate the actual cost to the study. Second, South African health worker wage scales were used to estimate the cost of CHW time if such an effort were conducted within the national health system.
The costs of fuel and maintenance for vehicles used by the CHWs to make their visits were extracted from the study's cost accounting system. The system produced monthly reports detailing fuel and maintenance costs for each study vehicle. Since the vehicles were also used for other tasks such as transportation of specimens to laboratories for analysis, only a percentage of the total vehicle costs were truly attributable to the CHW visits. This percentage was calculated as the relative proportion of time that each vehicle was used by the CHWs to make home visits.
Vehicle depreciation was estimated based on the initial purchase price of the vehicles, the expected useful life of each vehicle, and the assumption that the final resale value of the vehicle would be negligible. Straight line depreciation was employed to calculate the total depreciation for one year, using a 3% discount rate [24,25]. The same percentage of fuel and maintenance costs attributable to CHW visits was applied to the total vehicle depreciation cost to calculate vehicle depreciation attributable to CHW visits.
All costs are presented in 2003 South African Rand and do not include any form of tax since taxes represent transfer payments and are not real economic costs [24]. The exchange rate between 2003 South African Rand and 2003 US Dollars was 7.56 Rand per Dollar [26].
CHW visits
Since it was possible that multiple participants were visited on each CHW trip; that more than one CHW went on each trip; and that some participants received multiple CHW visits, we accounted for this in our estimates using data from the weekly CHW visit logs. To minimize potential bias of the number of CHW visits per participant due to censoring (e.g., additional CHW visits in early 2004 not counted for participants first visited in December of 2003), data on average number of CHW visits per participant was based solely on participants with scheduled appointments in the first half of 2003.
Cost per appointment
Based upon the total number of CHW home visits conducted and the total cost to carrying out these home visits, a cost per CHW home visit was calculated. Cost per woman screened was calculated by appointment type because the time since the previous visit differed by appointment type: 1) 5 months between the 1-month visit and the 6-month visit; 2) 6 months between the 6-month visit and the 12-month visit; 3) 12 months between the 12-month visit and the 24-month visit. To do this, the ratio of the number of CHW visits conducted for a particular appointment type to the total number of women attending that appointment type was calculated. Then, this ratio was multiplied by the cost per CHW visit to derive the marginal cost per woman for CHW home visits.
Ranges and sensitivity analyses
Because there was uncertainty in a number of parameters used to estimate total cost, we estimated an upper and lower bound for costs, reflecting a combination of differing assumptions regarding salary and percentage of vehicle costs attributable to CHW home visits. Because it was necessary to infer the appointment type for which CHW home visits were attempting to reduce loss to follow-up, we applied alternate assumptions to generate ranges of estimates of cost per woman screened.
Results
Effectiveness of CHW visits
Potential loss to follow-up differs by appointment type as does the success of CHW visits in preventing loss to follow-up. Figure 1 shows the distribution of the number of visits for participants with appointments in the first half of 2003 who returned for their appointments after CHW visits and for those who did not return. For those women who missed their scheduled appointment, the mean number of CHW visits was 1.98 for those with 6-month appointments, 2.06 for those with 12-month appointments, and 2.05 for those with 24-month appointments.
Figure 1 CHW Visits for Participants with Appointments in the First Half of 2003. The distribution of the number of CHW home visits for those women who eventually returned for their appointments (black bars) and those women who never returned for their appointments (white bars).
Figure 2 shows the percentage of women who returned without CHW visits, the percentage of women who returned after CHW visits, and those who did not return despite CHW visits. Without CHW visits, loss to follow-up for participants scheduled for 6 month, 12 month, and 24-month visits would have been 21%, 39%, and 50% respectively. With CHW visits, loss to follow-up was reduced to 6%, 10%, and 24% respectively.
Figure 2 Appointments Attended by Type and CHW Visit Status. The number and percentage of women who attended appointments without CHW home visits (green bars), women who attended appointments after CHW home visits (yellow bars), and women who never returned for their appointments (red bars) for each appointment type.
Quantities of resources used
CHW home visits involve groups of CHWs driving into Khayelitsha and surrounding township areas to visit study participants who have not returned to the clinic for their scheduled appointments. For security reasons, CHWs travel in groups of up to four, with larger groups generally going when informal settlements are visited.
In total, CHWs conducted approximately 3,200 visits in 530 trips during 2003. Means for staff per trip, participants visited per trip, time per trip, and time per participant visit were calculated along with their standard deviations from weekly CHW home visit logs (Table 1).
Table 1 Community Health Worker Visits to Participants
Estimate SD
CHWs per trip (trips = 457)* 2.82 0.85
Participants visited per trip (trips = 530)* 5.65 2.46
Time per trip (min) (trips = 386)* 111.32 52.60
Time per participant visited (min) (trips = 386)* 22.09 12.87
* The numbers of trips for these parameters differ because not all information was provided on all weekly report forms. To calculate ratios, only trips for which all information was available were included.
Primary observations of CHWs were conducted on several days to corroborate the information contained in the weekly logs. This observation found average visit time was 20.75 minutes (SD = 8.56, participants = 16). This matched the information from the reports (22.09 minutes, SD = 12.87, trips = 386). Additionally, it was found that approximately 70% of the participant visit time was spent driving between each participant's home.
On the 230 days per year that the project operates, its vehicles are used for three main purposes, staff transport, specimen transport, and CHW visits to participants. Approximately three hours per day are spent on staff and specimen transport. Approximately four hours and 20 minutes are spent on CHW visits per day with 70% of this time spent driving. Thus, 50% of the vehicles' costs are allocated to CHW visits.
Costs of resources used
Average monthly salaries for CHWs working for the KCCSP are approximately 4000 Rand, while average salaries for CHWs working in South Africa are approximately 2000 Rand per month.
The project operates four vehicles. The purchase price of each vehicle was approximately 40,000 Rand, excluding taxes. The mean and standard deviation of yearly fuel costs per vehicle are 9,111 Rand (SD = 1,735). The mean and standard deviation of yearly maintenance costs are 6,761 Rand (SD = 2,189). Using a 3% discount rate and five year, straight line depreciation, the yearly depreciation cost of each vehicle was 8,684 Rand. Table 2 presents the component costs attributable to CHW visits.
Table 2 Component Costs Attributable to CHW Visits (2003 Rand)
Estimate
CHW Time Cost 43,335
Fuel Cost 18,222
Maintenance Cost 13,522
Depreciation Cost 17,368
Total Cost 92,447
Cost per screened participant
The cost per CHW visit was 28.32 Rand. If the higher study CHW wage of 4,000 Rand per month was used and 80% of CHW visit time was spent driving, the estimate increases to 39.96 Rand per CHW visit. If an even lower, public sector wage of 1,500 Rand per month and 50% of CHW visit time was spent driving, the estimate decreases to 21.14 Rand per CHW visit.
To attribute these costs on a per-screened participant basis, it is necessary to examine the relative number of appointments and relative effectiveness of CHW visits at reducing loss to follow-up for 6, 12, and 24-month appointments. Table 3 shows the results of this analysis. Because of high success in CHW visits leading to participant return, 6-month per participant costs are 12.75 Rand. There is an increasing trend in per participant costs for the 12 and 24-month visits, with per participant cost equaling 24.92 Rand and 40.50 Rand respectively.
Table 3 CHW Visit Cost Attribution by Appointment Type (2003 Rand)
6-Month Visit 12-Month Visit 24-Month Visit
Participants Attending Appointments 861 1,635 743
CHW Visits Conducted 384 1,438 1,062
Visits per Participant 0.45 0.88 1.43
CHW Visit Cost per Participant (base case) 12.75 24.92 40.50
CHW Visit Cost per Participant (lower estimate) 9.51 18.60 30.23
CHW Visit Cost per Participant (upper estimate) 17.98 35.16 57.14
There was some ambiguity in the participant data as to which appointment type a particular CHW visit should be attributed. For example, for a participant with 6 and 12-month appointments within 2003 who missed both appointments and who received 4 CHW visits, it was not always clear which CHW visits led to attendance at each appointment. In the base case, we attributed the total number of CHW visits to both appointments since separate cost per woman estimates were calculated. However, this provides an overestimate of costs.
Two other alternatives were explored. In the first alternative, the earlier appointment was assumed to be the appointment for which all CHW home visits were necessary. In the second alternative, when such an ambiguity existed, the count of CHW visits was divided evenly between the two appointments.
Figure 3 shows the CHW home visit cost per woman using different inference rules to assign CHW home visits to the various appointment types. Per-woman screened costs ranged from 8.59–12.62 Rand, 20.65–24.89 Rand, and 39.42–40.45 Rand for 6, 12, and 24-month appointments respectively.
Figure 3 Cost per Woman Screened by Time since Previous Appointment, Sensitivity Analysis. The sensitivity analysis shown is of the cost per woman screened for CHW visits using alternative assumptions. The base case (green line and squares) attributes CHW visits to all missed appointments occurring during 2003. Alternative assumption 1 (blue line and triangles) attributes CHW visits to the first missed appointment occurring during 2003. Alternative assumption 2 (orange line and circles) evenly divides CHW visits between all missed appointments occurring during 2003.
Discussion
The objective of this study was to estimate the costs and effectiveness associated with an intervention to reduce the loss to follow-up after an initial cervical cancer screening visit in a South African screening program delivered to women of lower socio-economic status. We found that loss to follow-up was reduced significantly by community health worker home visits, and this effect was most pronounced when the intervention occurred in close proximity to the initial appointment. The costs of the CHW intervention were substantial when considered in the context of the total per-woman cost of cervical cancer screening. The CHW home visit costs were 8–15%, 15–29%, and 25–47% of the total per-woman screened cost for 6, 12, and 24-month visits [13].
Ultimately, the costs of cervical cancer screening including efforts to reduce loss to follow-up must be considered in relationship to the level of cervical cancer reduction. In an ethical study conducted in a reasonable time-frame, it would not be possible to measure actual reduction in cervical cancer incidence due to reduction in loss to follow-up. Thus, model-based studies are best suited to use our estimates to refine results from previous cost-effectiveness analyses of cervical cancer screening programs in developing countries. In addition, analysts conducting cost studies for short-term budgetary planning might wish to include these costs to reflect the full range of resources likely to be required in a new cervical cancer screening program.
This study has several limitations. First, we relied on data from a single year derived from sites in one peri-urban township outside of Cape Town. While the areas where participants live are spread over kilometers of land and in the winter rainy seasons, the unpaved roads of the informal settlements become difficult to navigate, the generalizability of these results to other parts of South Africa or, for that matter, to other developing countries is uncertain. Second, the data were derived from CHW home visits connected to a clinical research study. Possible selection biases may operate to make participants more or less likely to miss clinical appointments or to respond to efforts to reduce loss to follow up. Because the staff is better paid and likely better trained than average public sector CHWs, differences in competence and motivation level may also impact the success if this effort were to be replicated elsewhere. Third, other methods of reducing loss to follow-up such as telephone calls, letters, and specifically-targeted preventive educational sessions were not evaluated as alternatives to CHW visits. Because most participants reported not having telephones and because of difficulties in timely and accurate mail deliveries to all participants, we believe that these interventions were less appropriate for reducing loss to follow-up in this setting. An important question remains as to whether other potentially less costly and/or more effective methods exist to reduce loss to follow-up. Finally, the sustainability of this effort over many years and locations has not been tested.
To reflect the variability and uncertainty inherent in delivering cervical cancer screening services we have produced plausible ranges of costs and effectiveness for efforts to reduce loss to follow-up. Such estimates of the costs and effectiveness of reducing loss to follow-up should be considered as programs are planned and scaled-up to national population coverage levels.
Conclusion
In a South African cervical cancer screening study, loss to follow-up was reduced by CHW visits to women who had not attended their regularly scheduled appointments. The effectiveness of the CHW intervention was higher for appointments scheduled closer to initial screening visits. The total costs associated with CHW visits were appreciable, and the per-woman cost increased as the time between initial appointment and scheduled appointment increased. The cost-effectiveness of preventing loss to follow-up is an important consideration in planning national screening programs in resource poor settings
List of Abbreviations
CEA: cost-effectiveness analysis
CHW: community health worker
HPV: human papillomavirus
KCCSP: Khayelitsha Cervical Cancer Screening Program
VIA: visual inspection with acetic acid
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
JGF, LED, and SJG participated in the conception and design of the study. LED and TCW provided study materials or patients. JGF performed the collection and assembly of data as well as the analysis and interpretation. JGF drafted the article, and JGF, LED, MD, LK, and SJG provided critical revision of the article for important intellectual content.
Acknowledgements
The support and assistance of the entire staff of the Khayelitsha screening program made this research possible.
This study was supported in part by the Bill and Melinda Gates Foundation through grants to EngenderHealth; SJG is supported in part by the National Cancer Institute (R01-CA93435); JDG is supported in part by a National Science Foundation Graduate Research Fellowship.
==== Refs
Ferlay J Bray F Pisani P Parkin DM GLOBOCAN 2000: Cancer Incidence, Mortality and Prevalence Worldwide, Version 1.0 2001 IARC CancerBase No. 5. Lyon, IARCPress
Bosch FX Lorincz A Munoz N Meijer CJLM Shah KV The causal relationship between human papillomavirus and cervical cancer J Clin Pathol 2002 55 244 265 11919208
Ponten J Adami HO Bergstrom R Dillner J Friberg LG Strategies for global control of cervical cancer Int J Cancer 1995 60 1 26 7814140
Cuzick J Szarewski A Cubie H Hulman G Kitchener H Luesley D McGoogan E Menon U Terry G Edwards R Brooks C Desai M Gie C Ho L Jacobs I Pickles C Sasieni P Management of women who test positive for high-risk types of human papillomavirus: the HART study Lancet 2003 362 1871 6 14667741 10.1016/S0140-6736(03)14955-0
Saslow D Runowicz CD Solomon D Moscicki AB Smith RA Eyre HJ Cohen C American Cancer Society American Cancer Society guideline for the early detection of cervical neoplasia and cancer CA Cancer J Clin 2002 52 342 62 12469763
Sankaranarayanan R Budukh AM Rajkumar R Effective screening programmes for cervical cancer in low- and middle-income developing countries Bull World Health Organ 2001 79 954 62 11693978
Franco EL Primary screening of cervical cancer with human papillomavirus tests J Natl Cancer Inst Monogr 2003 31 89 96 12807951
Wright TC Jr Cervical cancer screening using visualization techniques J Natl Cancer Inst Monogr 2003 31 66 71 12807948
Denny L Kuhn L De Souza M Pollack AE Dupree W Wright TC Screen-and-Treat Approaches for Cervical Cancer Prevention in Low-Resource Settings JAMA 2005 294 2173 81 16264158 10.1001/jama.294.17.2173
Gaffikin L Blumenthal PD Emerson M Limpaphayom K Royal Thai College of Obstetricians and Gynaecologists (RTCOG)/JHPIEGO Corporation Cervical Cancer Prevention Group Safety, acceptability, and feasibility of a single-visit approach to cervical-cancer prevention in rural Thailand: a demonstration project Lancet 2003 361 814 20 12642047 10.1016/S0140-6736(03)13713-0
Denny L Kuhn L Risi L Richart RM Pollack A Lorincz A Two-stage cervical cancer screening: an alternative for resource-poor settings Am J Obstet Gynecol 2000 183 383 8 10942474 10.1067/mob.2000.105871
Mandelblatt JS Lawrence WF Gaffikin L Limpahayom KK Lumbiganon P Warakamin S Costs and benefits of different strategies to screen for cervical cancer in less-developed countries J Natl Cancer Inst 2002 94 1469 83 12359856
Goldie SJ Kuhn L Denny L Pollack A Wright TC Policy analysis of cervical cancer screening strategies in low-resource settings: clinical benefits and cost-effectiveness JAMA 2001 285 3107 15 11427139 10.1001/jama.285.24.3107
Goldie SJ Goldhaber-Fiebert JD Kim JJ Cost-Effectiveness of Cervical Cancer Screening in Kenya, India, Peru, South Africa, and Thailand Presented to the Society for Medical Decision Making 2004
Thinkhamrop J Lumbiganon P Jitpakdeebodin S Loss to follow-up of patients with abnormal Pap smear: magnitude and reasons J Med Assoc Thai 1998 81 862 5 9803086
Sankaranarayanan R Rajkumar R Theresa R Esmy PO Mahe C Bagyalakshmi KR Initial results from a randomized trial of cervical visual screening in rural south India Int J Cancer 2004 109 461 7 14961588 10.1002/ijc.11726
Cronje HS Screening for cervical cancer in developing countries Int J Gynaecol Obstet 2004 84 101 8 14871510 10.1016/j.ijgo.2003.09.009
Cronje HS Parham GP Cooreman BF de Beer A Divall P Bam RH A comparison of four screening methods for cervical neoplasia in a developing country Am J Obstet Gynecol 2003 188 395 400 12592246 10.1067/mob.2003.153
Cronje HS Trumpelmann MD Divall PD Scott LL Middlecote BD De Wet JI Cervical cytological services in the Orange Free State. Demographic characteristics S Afr Med J 1989 76 615 8 2595490
Doh AS Nkele NN Achu P Essimbi F Essame O Nkegoum B Visual inspection with acetic acid and cytology as screening methods for cervical lesions in Cameroon Int J Gynaecol Obstet 2005 89 167 73 15847890 10.1016/j.ijgo.2004.12.040
University of Zimbabwe/JHPIEGO Cervical Cancer Project Visual inspection with acetic acid for cervical cancer screening: test qualities in a primary care setting Lancet 1999 353 869 73 10093978 10.1016/S0140-6736(98)07033-0
Kuhn L Denny L Pollack A Lorincz A Richart RM Wright TC Human papillomavirus DNA testing for cervical cancer screening in low-resource settings J Natl Cancer Inst 2000 92 818 25 10814677 10.1093/jnci/92.10.818
Denny L Kuhn L Pollack A Wright TC Jr Direct visual inspection for cervical cancer screening: an analysis of factors influencing test performance Cancer 2002 94 1699 707 11920531 10.1002/cncr.10381
Michael Drummond Bernard O'Brien Greg Stoddart George Torrance Drummond MF (Editor) Methods for the Economic Evaluation of Health Care Programs 1997 2 Oxford; New York: Oxford University Press
Gold Marthe R Siegel Joanna E Russell Louise B Milton Weinstein C (eds) Cost-effectiveness in health and medicine 1996 Oxford; New York: Oxford University Press
EIU Country Data. Economist Intelligence Unit. London (Accessed: February 9, 2005).
|
16288646
|
PMC1308836
|
CC BY
|
2021-01-04 16:39:14
|
no
|
Cost Eff Resour Alloc. 2005 Nov 15; 3:11
|
utf-8
|
Cost Eff Resour Alloc
| 2,005 |
10.1186/1478-7547-3-11
|
oa_comm
|
==== Front
Environ HealthEnvironmental Health1476-069XBioMed Central London 1476-069X-4-261628007510.1186/1476-069X-4-26ResearchFertility in four regions spanning large contrasts in serum levels of widespread persistent organochlorines: a cross-sectional study Toft Gunnar [email protected] Anna [email protected] Aleksander [email protected] Ane Marie [email protected] Anna [email protected] Henning Sloth [email protected] Jan K [email protected] Valentina [email protected] Andery [email protected] Marcello [email protected] Gian Carlo [email protected]ørgensen Eva C [email protected] Lars [email protected] Jens Peter [email protected] [email protected] Department of Occupational Medicine, Aarhus University Hospital, Noerrebrogade 44, build. 2C, DK – 8000 Aarhus C, Denmark2 Department of Occupational and Environmental Medicine, Lund University Hospital, SE – 22185 Lund, Sweden3 Scanian Andrology Centre, Fertility Centre, Malmö University Hospital, SE – 205 02 Malmö, Sweden4 The Primary Health Care Clinic, P.O. Box 1001, DK-3900 Nuuk, Greenland5 Department of Environmental Toxicology, National Institute of Hygiene, Chocimska 24, PL – 00-791 Warsaw, Poland6 Problem Laboratory of Human Reproduction, Kharkiv State Medical University, Klochkovskaya Street 156-A, r. 14, 61145 Kharkiv, Ukraine7 Section of Toxicology and Biomedical Sciences, ENEA Casaccia, Via Anguillarese 301, I – 00060 Rome, Italy8 Laboratory di Genetica, Dip. di Science Agrarie, University of Modena and Reggio Emilia, Viale Kennedy 17, I – 42100 Reggio Emilia, Italy9 Unit of Environmental Biotechnology, Department of Environmental and Occupational Medicine, Institute of Public Health, Vennelyst Boulevard 6, Build. 260, University of Aarhus, DK-8000 Aarhus, Denmark2005 9 11 2005 4 26 26 11 8 2005 9 11 2005 Copyright © 2005 Toft et al; licensee BioMed Central Ltd.2005Toft et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Persistent organochlorine pollutants (POPs) may interfere with reproductive function but direct evidence in humans is very limited.
Methods
Fertility was examined in four regions with contrasting blood levels of POPs. Pregnant women and their partners in Warsaw (Poland), Kharkiv (Ukraine) and Greenland were consecutively enrolled during antenatal visits. Swedish fishermen and their spouses were recruited separately and independently of current pregnancy. Lipid adjusted serum concentrations of 2,2',4,4',5,5'-hexachlorobiphenyl (CB-153) and 1,1-dichloro-2,2-bis (p-chlorophenyl)-ethylene (DDE) were available for both partners. Time to pregnancy interviews were obtained among 2269 women and 798 men provided a semen sample.
Results
Inuits had high levels of both POP markers, Swedish fishermen were high in CB-153 but low in DDE, men from Kharkiv were high in DDE and low in CB-153 while men from Warsaw were low in CB-153 and had intermediate DDE levels. Compared to Warsaw couples, fecundability was reduced among couples from Kharkiv [adjusted fecundability ratio (FR) 0.64 (95% CI 0.5–0.8)] and elevated in Swedish fishermen families [FR 1.26 (95% CI 1.0–1.6)]. Adjusted geometric means of sperm counts and morphology did not differ between regions while sperm motility was higher in men living in Warsaw.
Conclusion
We observed regional differences in time to pregnancy and sperm motility that may be related to regional differences in POP blood levels, but other interpretations are also plausible. In particular, differences in access to safe contraception and in the prevalence of contraceptive failures are most likely to bias comparisons of time to pregnancy.
==== Body
Background
In 1993 it was proposed that widespread environmental man-made chemicals with agonistic or antagonistic effects on hormonal steroid receptors, in particular the estrogen receptor, might disturb the foetal development of the male gonads resulting in impaired multiplication of Sertoli cells leading to a permanent reduction of sperm counts [1,2]. Although experimental research offers some support to the hormone hypothesis, the epidemiological evidence on effects of xenobiotics mimicking hormones remains scarce. A recent review of 81 epidemiological studies linking indicators of prenatal serum levels of maternal estrogens with sperm density, hypospadias, cryptorchidism and testicular cancer concluded that – except for testicular cancer – there is no strong epidemiological evidence to indicate that prenatal exposure to estrogens are linked to disturbed development of the male reproductive organs [3,4]. It must be acknowledged, however, that epidemiological studies with adequate exposure assessment of environmental toxicants are very few [3,5].
In this paper we present initial findings from an ongoing, international study that was initiated in order to contribute additional evidence on the environmental hormone hypothesis by studies of persistent organochlorine pollutants (POPs). Polychlorinated biphenyls (PCBs) and the dichlorodiphenyl trichloroethane (DDT) metabolite 1,1-dichloro-2,2-bis (p-chlorophenyl)-ethylene (DDE) were selected as exposures of interest because of documented weak hormonal actions [6-15], widespread occurrence worldwide [16], existence of reliable and relatively inexpensive biomarkers suitable for large scale epidemiological studies [17], and the possibility to identify populations and groups with high contrast of exposure [18]. The objective of this paper was to examine whether the regional differences in organochlorine exposure levels are paralleled by differences in couple fertility and semen quality at the population level. In subsequent papers we examine associations between POP concentrations in serum and functional as well as biological indicators of fertility at the individual level [19].
Materials and methods
Study design
The basic study design combines four separate interview studies of time to pregnancy with four cross-sectional studies of semen quality using individual measurements of POP serum biomarkers to document exposure gradients between populations.
Recruitment of study populations
The target populations encompass pregnant women and their male spouses who had antenatal care visits from May 2002 through February 2004 at a large central hospital in Warsaw in Poland, at 3 hospitals and eight antenatal clinics in Kharkiv, Ukraine and at local hospitals in 19 municipalities and settlements throughout Greenland. With few exceptions the antenatal health programs cover all pregnant women in these localities. Altogether 3794 pregnant women and their spouses (if known) from these three localities were informed about the study and encouraged to list up when they, for the first time during the study period, attended the hospitals. Pregnant women were enlisted regardless of earlier reproductive history, parity or gestational age at the time of enrolment. In total 1710 pregnant women (45%) were included in the study (Table 1). The male spouses were consecutively encouraged to collect one semen sample until some 200 men at each site had agreed.
Table 1 Target populations and data collection in an international study of fertility.
Warsawa Kharkiva Greenlanda Swedenb Total
Eligible female target population, n 690 2478 665 1439 5272
Interviewed, n 472 640 598 559 2269
Participation rate, % 68 26 90 39 43
Contraceptive failures, % 19 48 6 21 26
Provided valid TTPc, n 376 307 520 519 1722
Stated TTP, n 369 301 455 513 1638
Calculated TTP, n 7 4 57 0 68
TTP in cyclesd, n 0 2 8 6 16
Questionnaire, mene, n 472 576 637 195 1880
Eligible men addressed for the semen study 690 640 256 2783 4369
Collected semen samplesf 198 208 201 191 798
Participation rate, % 29 33 79 7 18
Sperm motility, manual counting 190 208 200 188 786
Sperm motility, computer assisted 165 0 198 179 542
Sperm morphology 197 206 200 184 787
a) Consecutive pregnant woman and their spouses.
b) Retrospective studies of past pregnancies.
c) TTP were considered not valid if the woman was using birth control when becoming pregnant (47 Inuits, 47 Warsaw women, and 297 Kharkiv women), did not provide information on whether she was using birth control (n = 26), was multipara and had not menstruated since her previous pregnancy (n = 16), or had not provided information on whether she had menstruated since her previous pregnancy (n = 5).
d) Possibly censored measure.
e) The total number of individuals with questionnaire information on males. Data were obtained by interviewing the spouse if no male interview was performed.
f) All collected semen samples were analysed for sperm concentration and volume.
In addition to the pregnant couples we included fishermen spouses for a retrospective TTP study addressing the latest planned pregnancy, and fishermen for the cross-sectional semen study. The fishermen and their wives were enrolled independently in two separate steps from a cohort of fishermen living at the west or east coasts of Sweden, which originally was created for purposes of studies of health effects related to high and low level PCB exposure [20,21]. The local ethical committees representing all participating populations approved the study and all subjects signed an informed consent.
Data collection
At enrolment into the study the women took part in a TTP interview and within one week the men provided a fresh semen sample except for 116 of the men from Greenland who delivered the sample within one year. Both women and men also had a venous blood sample drawn. All interviews as well as sampling, processing, storage and shipment of blood and semen samples were undertaken according to uniform study protocols, questionnaires, and forms. The Swedish questionnaire was slightly different reflecting the population-based nature of this study but the questions establishing the TTP were the same. The data collection in each of the four regions is briefly summarised below.
Warsaw, Poland
Pregnant women and their spouses who either visited the obstetric outpatient clinic of the Gynaecological and Obstetric Hospital of the Warsaw School of Medicine or physicians at a collaborating hospital in the same city were informed about the project when attending antenatal classes at the hospitals. The coverage of the hospitals includes central areas of Warsaw as well as suburbs and rural areas close to the city. The antenatal classes typically included some 20–30 women. An obstetrician or midwife informed about the project and encouraged the women to consider participation in the study. Among those 690 women, who, after the classes in an individual talk with the project representative, explicitly accepted or denied the invitation, 472 were enlisted into the study and subsequently took part in a face-to-face TTP interview (participation rate 472/690, 68 %) and 198 of their male spouses agreed to provide a semen sample (participation rate 198/690, 29%). Fifteen midwives performed the TTP interviews when the women were on average 33 weeks pregnant (25–75 percentiles: 31.0–35.7) from September 2002 through March 2004. No data are available on those women and men who did not explicitly accept or decline participation.
Kharkiv, Ukraine
Altogether 2478 pregnant women and their spouses who visited one of eight antenatal clinics or three maternity hospitals in Kharkiv, Ukraine, were informed about the project and encouraged to participate. The population covered by the hospitals and clinics are partly living within the city of Kharkiv and partly in adjacent rural areas to the north east of the city. The women were individually contacted by one of 78 gynaecologists from the clinics and hospitals involved in the survey. Among the 2478 couples, 640 women (participation rate 640/2478, 26 %) and 208 of their male spouses (participation rate 208/640, 31%) were enrolled for the TTP and the semen study, respectively, from April 2003 through February 2004. The women were on average 24 weeks pregnant (25–75 percentiles: 12.1–33.6 weeks) weeks when the interview was performed. One of the reasons that a large number of pregnant women refused to participate was concern that the collection of a blood sample (altogether 35 ml whole blood) would imply a risk for the pregnancy and the baby – in particular when anaemia had been detected during pregnancy. Demographic and reproductive information was obtained from a sample of 605 of those women that declined participation in the study. The average age was slightly lower among non-participating women [22.8 (SD 2.4) versus 24.9 (SD 2.8) years], while the average number of children in the two groups was similar (1.1 versus 1.2 among those with at least one child).
Greenland
In total, 901 pregnant women from 15 municipalities (Aasiaat, Ilulissat, Kangaatsiaq, Maniitsoq, Nanortalik, Narsaq, Nuuk, Paamiut, Qaanaaq, Qaqortoq, Qasigiannguit, Qeqertarsuaq, Sisimiut, Uummannaq, Tasiilaq), including 4 settlements (Kulusuk, Kuummiut, Saattut and Ukkusissat) representing all regions of Greenland, were listed by the local midwife when visiting the local hospital or health clinic in June 2002 through May 2004. Two hundred and thirty six did not fulfil one or more eligibility criteria (108 women and 140 men were not born in Greenland and 43 were less than 18 years of age). One physician speaking the native language approached the remaining 665 and 598 (participation rate 598/665, 90%) provided a face-to-face TTP interview. The 67 eligible women that were not included in the study were out of range (32 women) or refused to participate (35 women). The age and parity distribution in participating and not participating eligible women were almost identical (data not shown). The women were on average 24 weeks pregnant (25–75 percentiles: 16.7–32.4 weeks) when interviewed. For the semen study 256 male partners were asked to participate. Thirty-five persons did not want to participate and 20 dropped out (18 could not be reached and 2 did not show up after 2 reminders) giving a total of 201semen samples (participation rate 201/256, 79%).
Sweden
Altogether 1439 fishermen wives, born 1945 or later, were contacted and 559 (participation rate 559/1439, 39 %) were enrolled for the TTP-study between December 2002 and March 2004. The age distribution among the participants (median 50 years, range 29–64) did not differ from that of those who declined participation (n = 691; median 51 years, range 27–58), or from non-respondents (n = 178; median 49 years, range 32–58). The men in the semen studies were recruited independently. In total 2783 men were asked in writing about their interest in taking part in a semen study and from 191 of these a semen and blood sample was collected during a visit to the residence in a mobile laboratory (participation rate 191/2783, 7%). To allow for comparisons with men recruited when their spouse were pregnant, we only included men that had ever fathered a child (79%).
The recruitment settings are not expected to affect the characteristics of the four study populations as such. For instance, none of the clinics or hospitals differentially received women with high-risk pregnancies. However, the use of an earlier cohort for the Swedish sample resulted in higher age of the fishermen and higher parity of the latest planned TTP study.
Interviews of female and male participants
We gathered information on TTP for the current (pregnancy based cohorts) or the latest planned pregnancy (the Swedish population based cohort) by face-to-face interviews with the women at the hospital or the residence or by telephone (Sweden). We used a structured interview questionnaire that was developed and validated in an earlier European Concerted Action [22]. After identifying those women that became pregnant without using birth control, TTP was assessed in three different ways. Firstly, the women were asked about their TTP using the following set of questions:
Leading up to this pregnancy, when was it that you started having sexual intercourse without using any birth control to prevent pregnancy?" Month:______ Year:______. We now call this the "STARTING TIME".
How long was it from that "STARTING TIME" until you became pregnant? (The date you became pregnant is the date you conceived) How long? Weeks:______ and/or Months:______ and/or Years:______
Of the women who provided information on TTP, 95% supplied information according to this method (Table 1). The questionnaire also contained a question on the date (month and year) when the couple stopped using birth control. Moreover, the date of the last menstrual period was established for all Swedish women as well as for all women from Greenland, Warsaw and Kharkiv who did not use any birth control, or who had menstruated since a previous pregnancy. If the woman had not provided a TTP according to the questions above, these two dates were used to calculate a TTP (4% of all TTPs; Table 1). If information about the day of the month was missing it was defined as the 15 th in the respective month. Finally, the women were asked about how many times they menstruated in the time period when they tried to become pregnant (0, 1, 2, 3 or more than 3 times). If no other measure of TTP was available, this possibly censored measure was used (1% of all TTPs; Table 1).
In addition the interview included questions on demographic and social factors, diet, lifestyle, medical history, job title, and occupational exposures. Information about time varying exposures as tobacco consumption and intake of alcoholic beverages was given with reference to the date when the couple started trying to become pregnant.
Similarly, all male partners that agreed to participate were interviewed to obtain information on lifestyle factors, urogenital disorders, occupational factors as well as questions about periods of abstinence and other issues relating to delivery of a semen sample. These data were used to describe reproductive characteristics of couples with men providing and not providing semen samples. The Swedish subsample was not included in these analyses because of limited overlap between female and male participants and the delayed sampling of semen relative to the latest planned pregnancy.
Both female and male questionnaires were translated to native language in the participating countries and back translated to English for correction of errors that occurred during the translation process. To minimize errors during typing in, all questionnaires were centrally typed in with inconsistencies in 1.7 % of the two sets of typing. When inconsistencies between the two sets of typing occurred the original data was looked up and the typed in data was corrected if necessary.
Measures and levels of exposure markers
Serum concentrations of 2,2',4,4',5,5'-hexachlorobiphenyl (CB-153) as a biomarker of PCB exposure and DDE as a marker of DDT exposure among 1445 women and 1172 male spouses or fishermen were analysed by gas chromatography mass spectrometry following solid phase extraction. Collection of samples and laboratory methods have been described in detail elsewhere [21,23]. CB-153 and DDE levels were adjusted for serum concentrations of triglycerides and cholesterol that were determined by enzymatic methods [18].
A comprehensive description of exposure profiles is given in [18]. In short, the median lipid adjusted serum concentrations of CB-153 were the highest among Inuit men (200 ng/g lipid), slightly lower in Swedish fishermen (190 ng/g lipid) and substantially lower among men from Kharkiv (44 ng/g lipid) and Warsaw (17 ng/g lipid). On the contrary, the median serum concentrations of p,p'-DDE was highest in Kharkiv (930 ng/g lipid) and lowest in Swedish fishermen (240 ng/g lipid) with Inuit and Polish men falling in between (560 and 530 ng/g lipid). Thus the Inuit men have rather high levels of both PCB and p,p'-DDE, the Swedish fishermen high PCB but low p,p'-DDE, and men from Kharkiv and Warsaw low PCB but rather high p,p'-DDE. Serum levels among women were in general lower but with a regional distribution that paralleled the distribution among the men (data not shown).
Collection and analysis of semen samples
Semen samples were collected by masturbation at the residence or in privacy in a room at the hospital. The subjects were asked to abstain from sexual activities for at least two days before collecting the sample, and to note the actual abstinence time. If collected at home, the sample was kept close to the body to maintain a temperature close to 37°C when transporting it to the laboratory immediately after collection. The samples were analysed for motility and concentration according to a manual for the project based on the WHO manual for basic semen analysis [24]. All samples were analysed by one researcher in each country and all semen analysers had been trained in a series of three workshops held before and during the sample collection at the Fertility Centre, Malmö University Hospital, Sweden. The variation among analysers was minimal [25] with a median inter-individual CV of 8.1% and 11.1%, respectively, for concentration and motility (grade a+b) assessments. All analyses of sperm motility were initiated within 95 minutes and 95 % of the analyses were initiated within 60 minutes after ejaculation. The morphology of the sperms from all the countries in this project were determined centrally by two technicians at the Fertility Centre, Malmö University Hospital, on Papanicolaou stained smears using the WHO 1999 criteria.
Statistical analyses
Assuming an average menstrual cycle length of one month, we estimated the probability of clinical pregnancy in a cycle among women not pregnant in the preceding cycle, conditionally that they did become pregnant. Since TTP was measured in discrete times (0, 1, 2, 3 n months), Fecundability Ratios (FR) and the 95% confidence intervals (95% CI) were calculated by Cox proportional hazard regression with handling of ties. The Fecundability Ratio (FR) estimates the fecundability in each region compared to the Warsaw region, which was chosen as reference because it was the region with the lowest POP exposure.
In order to evaluate possible fertility related selection bias we also compared TTP among men providing and not providing semen samples in the three pregnancy based populations where the semen studies were performed among subsets of all couples enrolled into the study (Warsaw, Kharkiv and Inuits). In all analyses couples that became pregnant in spite of contraception or with otherwise undefined or missing TTP values (n= 537) were excluded. TTP values above 12 months (n = 253) were censored to account for medical treatment for infertility, which is usually not started until after one year of infertility. We also performed sensitivity analyses that (i) used censoring after 18 and 24 months, (ii) included couples that became pregnant in spite of use of contraception (TTP assigned 0), (iii) included first parity pregnancies only, (iv) and that included couples than discontinued non-oral contraception only. All analyses were adjusted for established determinants of fecundability, namely female age and female smoking – while parity, other female lifestyle factors, current employment, reproductive disease history, frequency of intercourse and use of oral contraceptive were included in the models if the risk estimate changed more than 10% in any population.
The distribution of semen characteristics in the four populations was compared overall and in pairs by analysis of variance (ANOVA) and multiple linear regression. Sperm concentration, total sperm count, percentages of sperms with normal morphology, age and abstinence time were transformed by the natural logarithmic function, which improved normality and homogeneity of variance. The percentage of motile sperms was not transformed. The mean values for sperm concentration and total sperm count on the logarithmic scale in each region were adjusted for duration of sexual abstinence before delivery of the semen sample in all analyses (PROC GLM, the LSMEANS options). The model fit was evaluated by visual inspection of residual plots. Geometric mean values and their confidence limits were obtained by back transformation. Other potential confounding determinants of semen characteristics as age, spillage, body mass index, fever last three months, season of sampling, urogenital infections or surgery were included in the models if the regression coefficients were changed by more than 10% for any population [26]. Motility analysis was restricted to the 95 % of samples for which the analysis was initiated within 60 minutes after collection.
All analyses were performed using SAS 9.13 software [27]. The term statistically significant is used to denote a p-value less than 0.05.
Results
Data related to creation of cohorts and participation at various levels are displayed in Table 1. The highest rate of participation in TTP interviews as well as in the semen studies were encountered in Greenland while participation was substantially lower in the other study groups.
Table 2 and 3 presents major differences between the regions with respect to demographic characteristics, lifestyle factors, reproductive behaviour and prevalence of urogenital infections. Thus use and type of contraception was strongly divergent in the four regions. In the Kharkiv cohort, TTP was not defined in almost half of the couples because pregnancies were not planned and occurred in spite of contraception, mostly less safe contraceptive methods (Table 3). However, within regions male and female characteristics were much alike among those couples where the man provided a semen sample compared to those that did not deliver a sample (data not shown).
Table 2 Characteristics of women providing time to pregnancy interviews (n = 2269) and men providing semen samples (n = 798).
Warsaw Kharkiv Greenland Sweden
Women
N = 472 Men
N = 198 Women
N = 640 Men
N = 208 Women
N = 598 Men
N = 201 Women
N = 559 Men
N = 191
Age, mean (SD) 29 (4) 30 (4) 25 (5) 26 (5) 27 (6) 31 (7) 29 (5) 47 (10)
Employed, % 89 97 61 75 51 83 70 95
Body mass index, kg/m2, mean (SD) 21 (3) 26 (3) 22 (3) 24 (3) 25 (4) 26 (5) 25 (4) 27 (3)
Current smoking, % 19 27 22 66 73 72 31 23
Alcoholic beverage consumption, drinks/ week, mean (SD) 2 (3) 6 (6) 1 (8) 3 (2) 2 (6) 9 (14) 1 (3) n.a
Coffee, cups/week, mean (SD) 1 (1) 2 (2) 2 (1) 2 (2) 2 (3) 5 (5) 3 (3) n.a
Urogenital infectionsa, % 14 5 13 5 85 83 27 21
Urogenital disordersb, % 21 3 12 3 7 1 44 2
Primi parity, % 91 - 79 - 31 - 4 -
Intercourse daily, % 18 - 19 - 42 - 4 -
a) Chlamydia infection, gonorrhoea, other sexually transmitted diseases or for women: Pelvic infections after delivery; for men: epididymitis or mumps in adulthood.
b) For women: Ovarian cysts, fibroids, myomas, endometriosis, cancer therapy and surgery on uterus, salpinges or ovaries. For men: Treatment for retracted testis, varicocele, testicular torsion, and testicular cancer.
Table 3 Type of contraception by categories of time to pregnancy (TTP). Row percentages.
The pill Coil or implant Condom or diaphragm Withdrawal, safe periods, other No contra-ception Missing data
TTP defineda
Warsaw n = 376 25 1 31 28 7 8
Kharkiv n = 307 3 1 16 40 16 24
Greenland n = 520 39 16 5 4 27 9
Sweden n = 519 25 29 25 9 0 10
Contraceptive failuresb
Warsaw n = 89 6 2 41 52 0 0
Kharkiv n = 310 5 3 35 56 0 1
Greenland n = 38 40 8 29 24 0 0
Swedenc n = 150d 29 10 15 3 0 42
Missing TTP datae
Warsaw n = 7 0 0 43 14 29 14
Kharkiv n = 23 8 4 4 39 4 1
Greenland n = 40 25 5 5 5 18 43
Sweden n = 40 13 3 13 3 0 70
a) Time from a well defined start of unprotected sexual intercourse until last menstrual period
b) Became pregnant in spite of use of contraception
c) Frequencies based upon first unplanned pregnancies
d) The number of unplanned latest pregnancies
e) Including women that became pregnant following delivery without recurrence of menstrual bleeding (n = 16)
The fecundability among couples in Kharkiv was lower than among couples in Greenland and Warsaw and among fishermen families (Table 4). This difference became even more pronounced when the risk estimates were adjusted for female age, female smoking and several other known or suspected risk factors. Fishermen families exhibited increased fecundability although adjustment for potential confounders weakened the association to borderline significance. Similar findings were obtained when TTP values were censored after 18 and 24 months rather than after 12 months, in analyses only including couples not using oral contraception and in analyses only including the first pregnancies except that the latter analysis revealed a reduced fecundability among fishermen's families (Table 4, the foot note). Inclusion of couples that became pregnant in spite of contraception by assigning these couples a TTP value of 0 resulted in weaker differences between regions.
Table 4 Fecundability ratiosa (FR) for couples in Kharkiv, Greenland and Sweden in comparison with couples in Warsaw.
Number pregnant (row percentage) FR crude FR adjustedb 95% CI
0 – 3 months 4 – 6 months 6 – 12 months > 12 months
Warsaw 208 (55) 38 (10) 60 (16) 70 (19) 1.00 1.00 -
Kharkiv 147 (48) 37 (12) 39 (13) 84 (27) 0.80 0.64 0.53 – 0.79
Greenland 295 (57) 61 (12) 85 (46) 80 (15) 1.07 1.00 0.80 – 1.27
Sweden 355 (68) 75 (15) 36 (7) 53 (10) 1.35 1.26 1.01 – 1.57
a) Adjustment for female age (continuous), female current smoking (yes/no), primiparity (yes/no), daily sexual intercourse (yes/no) and current employment (yes/no). All covariates except current employment contributed significantly to the model. Female urogenital disease or infection, low and high body mass index and use of oral contraceptives did not change at least one of the risk estimates with at least 10%.
b) The corresponding age and smoking adjusted FR among nulliparous women were (i) Kharkiv: 0.59 (95% CI, N = 503) (ii) Greenland: 0.90 (95% CI 0.68–1.19, N = 187) and Sweden FR 0.90, (95% CI 0.73 – 1.11), N = 429]. The Swedish risk estimate based upon the first planned pregnancy.
Ranking of regions according to median serum levels of DDE and couple fecundability resulted in identical successions while couple fertility at the regional level seemed unrelated to the average CB-153 levels.
The fecundability decreased with increasing age, was reduced among female smokers, in primi-parae, in less sexually active couples and in women reporting urogenital disorders. Fecundability was slightly lower among couples providing semen samples compared to those not providing samples in all regions, but the difference was not statistically significant in any of the three regions or overall (Table 5).
Table 5 Fecundability ratiosa (FR) for couples providing semen samples in comparison with couples not providing semenb.
Number pregnant (row percentage) FR crude FR adjusted 95% CI
0 – 3 months 4 – 6 months 6 – 12 months > 12 months
Warsaw
- sample 123 (56) 23 (10) 37 (17) 38 (17) 1.00 1.00 -
+sample 85 (55) 15 (10) 23 (15) 32 (20) 0.91 0.90 0.7 – 1.1
Kharkiv
- sample 103 (49) 27 (13) 27 (13) 53 (25) 1.00 1.00 -
+sample 44 (46) 10 (10) 12 (12) 31 (32) 0.87 1.02 0.8 – 1.4
Greenland
- sample 205 (60) 33 (10) 50 (15) 53 (16) 1.00 1.00 -
+sample 89 (50) 28 (15) 35 (20) 27 (15) 0.89 0.88 0.7 – 1.1
All
- sample 431 (56) 83 (11) 114 (15) 144 (19) 1.00 1.00 -
+sample 218 (51) 53 (12) 70 (16) 90 (21) 0.89 0.91 0.80 – 1.04
a) Adjustment for region (only the overall estimate), female age (continuous) and current female smoking (yes/no).
b) Sweden not included (TTP and semen studies with marginal overlap).
The crude sperm concentration, the total sperm count, other semen characteristics as well as potential confounding factors recorded at the sampling time are presented in Table 6 for each study population and Table 7 reports crude and abstinence time adjusted geometric mean values with their 95% confidence limits as well as pair-wise comparisons between each region and the reference region (Warsaw).
Table 6 Unadjusted semen characteristics by region.
Warsaw
N = 198 Kharkiv
N = 208 Greenland
N = 201 Sweden
N = 191
Sperm concentration × 106/ml
mean (SD) 88 (80) 75 (61) 72 (61) 57 (44)
median (p5 – p95) 64 (7–258) 59 (10–193) 53 (11–178) 49 (9–166)
Total sperm count × 106
mean (SD) 343 (374) 255 (238) 245 (241) 182 (159)
median (p5 – p95) 197 (19–1071) 181 (24–746) 186 (32–667) 133 (15–500)
Volume ml, mean (SD) 3.8 (1.8) 3.5 (1.9) 3.5 (1.7) 3.3 (1.7)
A+B motile, % (SD) 60 (20) 54 (22) 55 (19) 57 (21)
CASA, motile, % (SD) 42 (22) not performed 50 (21) 32 (21)
Normal morphology, % (SD) 6.7 (3.8) 7.3 (4.2) 6.9 (3.7) 8.2 (4.5)
Abstinence period a, days,
mean (SD) 7.7 (9.5) 3.9 (2.0) 3.5 (3.1) 3.7 (2.6)
median (p5 – p95) 4.0 (1.0–30.0) 3.0 (1.5–7.0) 3.0 (0.5–7.0) 3.0 (1.0–8.0)
Time to analysis, minutes, mean (SD) 52 (8) 37 (12) 35 (18) 44 (12)
Season for sperm collection, %
Spring 19 26 32 34
Summer 10 22 54 35
Fall 21 24 14 31
Winter 50 28 0 0
Fever last three months, % 9 10 13 3
Pain at urination last three months, % 2 19 3 0
Infections, vaccines or surgery last six months, % 26 72 12 8
Medicine or natural medicine use last six months, % 33 68 32 33
Spillage of semen, % 6 17 11 16
a) Abstinence time only included from subjects reporting abstinence time <60 days.
Table 7 Geometric mean values and the 95% confidence interval for semen characteristics by region.
Warsaw Kharkiv Greenland Swedena P overall, crude/ adjustedb
Sperm concentration, mill/ml 51 (44;58) 55 (49;61) 57 (50;64) 45 (39;52) 0.04/0.08
Urogenital infections 23 (14; 40) 50 (30;82) 53 (47; 60) 44 (33;58) 0.22/0.02
No urogenital infection 54 (47;62) 56 (50;64) 62 (46; 83) 47 (40;54) 0.07/0.19
Sperm countc, millions 164 (142;188) 179 (155;206) 180 (156;207) 146 (123;173) 0.08/0.22
Motile spermatozoad, % 60 (57; 63) 54 (52; 57) 55 (52; 58) 56 (53; 60) 0.02/0.02
Normal morphology, % 6.7 (6.2;7.3) 7.1 (6.5;7.6) 7.1 (6.5;7.6) 7.9 (7.2;8.7) 0.07/0.07
Figures in bold indicate p < 0.05 in pair-wise comparisons with the Warsaw sample as reference (Dunnett method).
a) Men that have ever fathered a child.
b) Period of abstinence, count values only.|
c) Restricted to men reporting no spillage.
d) Arithmetical mean values.
The lowest adjusted sperm concentration was found in the Swedish fishermen and the adjusted sperm concentrations differed only slightly among men from Warsaw, Kharkiv and Greenland.
Potential confounders related to sampling and processing of the specimens (season, recent fewer and spillage) changed the effect on regression coefficients of sperm concentration less than 10% in all study groups and was therefore not included in the final model. Among other possible confounders only urogenital infections changed the estimate more than 10%, but only in Greenland, where urogenital infections are much more prevalent (83%) than in the other regions (Swedish fishermen 21% and Warsaw and Kharkiv 5%).
The sperm concentration, stratified by urogenital diseases is presented in Table 7, indicating a reduction in sperm concentration among men that reported urogenital diseases in all of the four study populations with a weighted average reduction of 17.5%. However, only in the samples from Warsaw the difference reached statistical significance. The very low sperm count among men from Warsaw with urogenital infections should be interpreted with caution since this number is based on only 9 subjects.
The distribution of total sperm count (concentration multiplied by volume) within and between regions paralleled sperm concentration. We did not observe obvious differences between regions of crude values (p = 0.08), sperm counts adjusted for trivial factors as period of abstinence (p = 0.22) or sperm counts adjusted for urogenital infections (p = 0.13).
The sperm motility (manual counting) differed among countries (p = 0.02), with the highest motility observed in semen samples from Warsaw and the lowest on semen samples from Kharkiv. The adjusted geometric mean percentage of men with morphological normal sperm differed up to 15% among regions with lowest values in Warsaw and highest values in Sweden (Table 7). Both for motility and percentage of normal sperms none of the tested potential confounders or explanatory covariates changed the estimate more than 10%.
Discussion
Comparison of health outcomes in regions with highly contrasting exposure levels may provide clues pointing to environmental causes of disease otherwise impossible to detect within populations [18,28]. This international study of fertility includes populations with body burdens of POPs ranking highest in the world and POP contrasts between regions reaching more than one order of magnitude. We observed regional differences in couple fecundability in terms of higher fecundability among Swedish fishermen's families that are characterised by a high PCB but low DDE blood levels and low fecundability in the Kharkiv region characterised by low PCB but high DDE levels. Thus fecundability was related to the average population blood levels of DDE but not PCB. On the contrary, semen quality was remarkable stable across regions except that sperm motility was increased in the Warsaw region having low PCB levels and medium DDE levels.
From these data it is tempting to speculate that persistent environmental pollutants might contribute to the regional variation of fertility. Reproductive function is regulated by hormonal signalling. Several in-vitro and ex-vivo studies indicate that POPs weakly interact with steroid receptors [29-32] and animal studies have reported effects on reproductive organs following administration of both various PCB congeners and mixtures [33,34] and DDT metabolites [35]. Few US and Swedish studies on POP related effects on TTP have been conflicting – some reporting effects and others not [3], but one study addressing partners of male DDT applicators found impaired fecundity [36]. Other compounds or mixtures of environmental toxicants, as for instance phthalates, might be of significance too. CB-153 and DDE are hardly reflecting the total exposure to hormonal active xenobiotics. Thus men living in Warsaw have relatively low levels of CB-153 and DDE but could have higher exposure to other compounds, which have not been considered in the present study. In a recent paper, Hauser et al claims a synergistic effect between PCB and phthalates on semen quality [37], although this finding was not corroborated in a Swedish study of military conscripts [38]. However, at present we have no data to indicate that people in Warsaw are more or less exposed to the wide range of xenobiotics that potentially could interfere with.
Inuits had high-level exposure to both POP markers but yet the fecundability and semen quality was not obviously reduced. It has previously been suggested that the putative negative effect of consumption of POP contaminated seafood (which is common among both the Swedish fishermen's families and the Inuits) may be outweighed by the positive effects of other constituents in this type of food, such as antioxidants and polyunsaturated fatty acids [39,40]. Oxidative stress in the genital tract may play a pivotal role for impaired reproductive function [41]. Excess free radical generation in spermatozoa is related to loss of motility and fertilizing potential because of peroxidation of unsaturated fatty acids in the sperm plasma membrane [41]. Furthermore, low sperm production and poor semen quality are consistent features of Se-deficient animals [42]. A higher dietary supply of antioxidants as selenium, a range of anti-oxidative vitamins as well as polyunsaturated fatty acids through seafood may outweigh deleterious effects of POPs and other toxicants that also are conveyed through seafood.
However, methodological considerations offer alternative explanations for the regional variation of fecundability and possibly some of the minor differences that were seen in semen quality. The number of contraceptive failures was much higher in the Kharkiv population. Since the TTP is not defined for unplanned pregnancies, the most fertile couples in these populations are excluded from the analyses. Thus, it seems likely that the low fecundability in the Kharkiv cohort is biased because of selection of less fertile couples. The sensitivity analysis that also included couples with contraceptive failures supported this interpretation since differences between regions became less pronounced.
The slightly higher fecundability in the Swedish sample may be explained by selection working in the opposite direction. The TTP analyses of the Swedish sample was based upon the latest planned pregnancy and thus includes the majority of assumed highly fertile couples that at other times experienced contraceptive failures – a subgroup of some 20% (Table 2). While a number of sensitivity analyses in general corroborated the reported findings, restriction to first pregnancies did reveal a reduced fecundability among the Swedish couples. Thus the observed differences in fecundability may be artefacts of design and analysis rather than effects of POP exposure or other dietary ingredients.
Time to pregnancy studies of time trends and regional differences are probably exceptionally susceptible to bias that may be difficult to identify [22,43]. Although the recall of the time taken to conceive is expected to be highly valid in the pregnancy based studies, across region differences in the average weeks (23 to 33) the women were pregnant at enrolment could cause biased comparisons of TTP if pregnancies that were terminated by a spontaneous abortion were not included on equal terms in the four regions. Earlier studies indicate links between subfecundity and spontaneous abortion [44]. Nevertheless, we believe this is a minor problem in this study because more than 86% of women were enrolled after the 12th week of pregnancy and thus is at low risk for abortion. However, if exposures in addition to subfertility first of all causes early abortions out study would not pick up such effects.
The lowest sperm concentration was found in the Swedish fishermen population with high CB-153 exposure but low DDE exposure while Inuit men, high in both CB-153 and DDE, had sperm counts close to men from Warsaw and Kharkiv – both regions low in CB-153 but high or very high in DDE. The apparent low sperm count among fishermen could reflect a considerably higher age among the fishermen. Age was negatively but not statistically significant related to sperm count (p = 0.29), but inclusion of age in the model did maximally change the estimate 7 %. A more likely explanation is that the Swedish sample includes more subfertile men than the other samples where spouses of currently pregnant women were recruited. Although we only included fishermen that previously had fathered a child, the most subfertile men taking several years to achieve a pregnancy is expected to be included to a higher extent among the elderly Swedish fishermen than among the rather young men from the other study groups [25,45].
The period of sexual abstinence was lowest in Inuit men. When adjusting for abstinence the Inuits' sperm concentration and count were the highest – indicating that high levels of CB-153 and DDE are not associated with severely impaired spermatogenesis.
While the present study provides little evidence that populations with high PCB and/or DDT exposure experiences reduced sperm counts or more morphologically abnormal sperms, there are indications that sperm motility could be affected. The adjusted sperm motility was higher in the population from Warsaw with low serum level of CB-153 and moderate levels of DDE. The manual assessment of sperm motility has often been demonstrated to be highly variable [46] and bias due to differences between the four technicians that performed manual motility counting at the four study sites is a concern. However, quality control workshops performed during the study revealed an inter-individual variation in motility assessment of only 11 % among the four technicians without any indication of systematic difference between the Polish technician and the others. Thus, we do not believe that difference in laboratory methods explain the findings. Moreover, findings in this study are consistent with two independent studies based upon exposure contrasts within the Swedish population [21,23]. We also performed computer-assisted motility counting in three of the four countries (including Poland) but, unfortunately, the variability between countries in performing the analysis was too large to allow for between region comparisons.
The percentage of sperms with normal morphology did not differ significantly between countries. All morphological smears were sent to one central laboratory for staining and assessment of sperm abnormalities by two technicians. Therefore, the systematic measurement error should be minimal on this outcome, although also this outcome is known to have a high degree of intra- and inter-observer variability [46].
Cross sectional studies of semen quality suffer most often of low participation rates that may bias the internal validity of a study. In occupational semen studies, subfertility seems to be a stronger motivation for participation among referents than among exposed workers, who may have an interest to have potential harmful exposures documented. Such a differential selection may, however, be less important in environmental studies where participants may know little or nothing about their individual exposure level. In this study, the fecundability was – as expected – lower among those men that provided a semen sample than those that did not, but differences were small and – more importantly – selection was not related to levels of exposure as indicated by the study group. Moreover, the distribution of lifestyle factors, occupational exposure, and reproductive disease history among men providing semen sample and non-providers demonstrated only small differences within countries. Thus, low participation rates are not expected to bias the relation between POP exposure and semen characteristics. In Sweden, only men who agreed to donate a semen sample were interviewed making us unable to draw the same conclusions for this population. As far as the female characteristics are concerned, almost all of the listed factors differed between the four countries, but only minor differences were seen between those women whose men provided a semen sample and all other women.
Although all studies were set-up and carried out according to an agreed uniform research protocol, the marked across region differences in demographic factors as age, lifestyle factors as smoking, urogenital infections, contraceptive methods, reproductive behaviour, periods of sexual abstinence and season of sample collection call for an adequate control for potential confounding by these factors in the between regions analyses. We choose the change-in-estimate method to keep factors in the models regardless of p-values (38) and known strong determinants as age and period of abstinence were compulsory in all models. Even some residual confounding cannot be ruled out, the main threat to the internal validity of this study is probably strong selective forces that cannot be adjusted for in the analyses.
Conclusion
In this large epidemiological study of fecundability and semen quality in regions spanning more than 10 fold differences in median serum concentrations of CB-153 and DDE, we observed regional variations in couple fecundability that are compatible with effects of DDE, but the regional variation can also be due to differences between regions with respect to recruitment, reproductive disorders, sexual behavior and use of contraception. We found no evidence that men living in areas with high POP exposure levels have low sperm counts or more abnormal sperm but men living in an area with rather low POP exposure had higher sperm motility. Subsequent work investigating the relation between individual measures of exposure and a range of fertility outcomes will hopefully provide additional insight. In addition to TTP and semen quality these studies will also include indicators of sperm chromatin structure, DNA damage and apoptosis, sexual hormones, X/Y chromosome ratio in sperm and epididymal and accessory gland function. Fertility-related selection to the semen studies with rather low participation rates seems not to be a major concern for the interpretation of our findings.
Abbreviations
ANOVA Analysis Of Variance
CB-153 2,2',4,4',5,5'-hexachlorobiphenyl
CI Confidence Interval
CV Coefficient of variation
DDE 1,1-dichloro-2,2-bis (p-chlorophenyl)-ethylene
DDT Dichlorodiphenyl Trichloroethane
FR Fecundability Ratio
GLM General linear model
PCB Polychlorinated biphenyl
POP Persistent Organochlorine Pollutant
TTP Time to Pregnancy
WHO World Health Organisation
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
JPB, AG and LH conceived and designed the project. JPB was main responsible for raising funding for the project. JPB and GT coordinated the execution of the project. ARH, GT, HSP, JKL, and VZ have been responsible for collecting all blood samples and for obtaining the interview data. GT had main responsibility for creating the joint database. GT and JPB performed the statistical analyses. GT and JPB drafted the manuscript. All authors participated in the design of the study, commented on the draft, and have read and approved the final manuscript.
Acknowledgements
The INUENDO Project: Biopersistent organochlorines in diet and human fertility. Epidemiological studies of TTP and semen quality in Inuit and European populations, is supported by The European Commission to the 5th Framework Programme Quality of Life and Management of living resources, Key action four on environment and health, (Contract no. QLK4-CT-2001-00202), running 01.01.02-30.06.05. . The Ukrainian part of the study was supported by a grant from INTAS (Contract no. 2001 2205). Grants were also received from Aarhus University. The Swedish part of the study has also been funded by the Swedish Medical Research Council and the Swedish Research Council for Environment, Agricultural sciences and Spatial Planning.
==== Refs
Sharpe RM Skakkebæk NE Are oestrogens involved in falling sperm counts and disorders of the male reproductive tract? The Lancet 1993 341 1392 1395 8098802 10.1016/0140-6736(93)90953-E
Sharpe RM Irvine DS How strong is the evidence of a link between environmental chemicals and adverse effects on human reproductive health? BMJ 2004 328 447 451 14976101 10.1136/bmj.328.7437.447
Toft G Hagmar L Giwercman A Bonde JP Epidemiological evidence on reproductive effects of persistent organochlorines in humans Reprod Toxicol 2004 19 5 26 15336708 10.1016/j.reprotox.2004.05.006
Storgaard L Bonde JP Olsen J Male reproductive disorders in humans and prenatal indicators of estrogen exposure A review of published epidemiological studies Reprod Toxicol 2005 . 16005180
Vidaeff AC Sever LE In utero exposure to environmental estrogens and male reproductive health: a systematic review of biological and epidemiologic evidence Reprod Toxicol 2005 20 5 20 15808781 10.1016/j.reprotox.2004.12.015
Bonefeld-Jorgensen EC Andersen HR Rasmussen TH Vinggaard AM Effect of highly bioaccumulated polychlorinated biphenyl congeners on estrogen and androgen receptor activity Toxicology 2001 158 141 153 11275356 10.1016/S0300-483X(00)00368-1
Connor K Ramamoorthy K Moore M Mustain M Chen I Safe S Zacharewski T Gillesby B Joyeux A Balaguer P Hydroxylated polychlorinated biphenyls (PCBs) as estrogens and antiestrogens: structure-activity relationships Toxicol Appl Pharmacol 1997 145 111 123 9221830 10.1006/taap.1997.8169
Fielden MR Chen I Chittim B Safe SH Zacharewski TR Examination of the estrogenicity of 2,4,6,2',6'-pentachlorobiphenyl (PCB 104), its hydroxylated metabolite 2,4,6,2',6'-pentachloro-4-biphenylol (HO-PCB 104), and a further chlorinated derivative, 2,4,6,2',4',6'-hexachlorobiphenyl (PCB 155) Environ Health Perspect 1997 105 1238 1248 9370515
Moore M Mustain M Daniel K Chen I Safe S Zacharewski T Gillesby B Joyeux A Balaguer P Antiestrogenic activity of hydroxylated polychlorinated biphenyl congeners identified in human serum Toxicol Appl Pharmacol 1997 142 160 168 9007045 10.1006/taap.1996.8022
Nesaretnam K Corcoran D Dils RR Darbre P 3,4,3',4'-Tetrachlorobiphenyl acts as an estrogen in vitro and in vivo Mol Endocrinol 1996 10 923 936 8843409 10.1210/me.10.8.923
Nesaretnam K Darbre P 3,5,3',5'-tetrachlorobiphenyl is a weak oestrogen agonist in vitro and in vivo J Steroid Biochem Mol Biol 1997 62 409 418 9449244 10.1016/S0960-0760(97)00062-9
O'Connor JC Frame SR Davis LG Cook JC Detection of the environmental antiandrogen p,p-DDE in CD and long-evans rats using a tier I screening battery and a Hershberger assay Toxicol Sci 1999 51 44 53 10496676 10.1093/toxsci/51.1.44
Kelce WR Lambright CR Gray LEJ Roberts KP Vinclozolin and p,p'-DDE alter androgen-dependent gene expression: in vivo confirmation of an androgen receptor-mediated mechanism Toxicol Appl Pharmacol 1997 142 192 200 9007049 10.1006/taap.1996.7966
Grunfeld HT Bonefeld-Jorgensen EC Effect of in vitro estrogenic pesticides on human oestrogen receptor alpha and beta mRNA levels Toxicol Lett 2004 151 467 480 15261991 10.1016/j.toxlet.2004.03.021
Rattenborg T Gjermandsen I Bonefeld-Jorgensen EC Inhibition of E2-induced expression of BRCA1 by persistent organochlorines Breast Cancer Res 2002 4 R12 12473173 10.1186/bcr461
Bignert A Olsson M Persson W Jensen S Zakrisson S Litzen K Eriksson U Haggberg L Alsberg T Temporal trends of organochlorines in Northern Europe, 1967-1995. Relation to global fractionation, leakage from sediments and international measures Environ Pollut 1998 99 177 198 15093312 10.1016/S0269-7491(97)00191-7
Grimvall E Rylander L Nilsson-Ehle P Nilsson U Stromberg U Hagmar L Ostman C Monitoring of polychlorinated biphenyls in human blood plasma: methodological developments and influence of age, lactation, and fish consumption Arch Environ Contam Toxicol 1997 32 329 336 9096084 10.1007/s002449900193
Jonsson B Rylander L Lindh C Rignell-Hydbom A Giwercman A Toft G Pedersen HS Ludwicki JK Zvyezday V Spano M Bizarro D E BJ Manicardi GC Bonde JP Hagmar L Inter-population variations in concentrations, determinants of and correlations between 2,2',4,4',5,5'-hexachlorobiphenyl (CB-153) and 1,1-dichloro-2,2-bis (p-chlorophenyl)-ethylene (p,p'-DDE): a cross-sectional study of 3161 men and women from Inuit and European populations. Environmental Health Perspectives 2005 4 27
Spano M Toft G Hagmar L Eleuteri P Rescia M Rignell-Hydbom A Tyrkiel E Zvyezday V Bonde JP Exposure to PCB and p,p'-DDE in European and Inuit populations: impact on human sperm chromatin integrity Hum Reprod 2005 . 16223788
Rylander L Stromberg U Hagmar L Decreased birthweight among infants born to women with a high dietary intake of fish contaminated with persistent organochlorine compounds Scand J Work Environ Health 1995 21 368 375 8571093
Rignell-Hydbom A Rylander L Giwercman A Jonsson BA Nilsson-Ehle P Hagmar L Exposure to CB-153 and p,p'-DDE and male reproductive function Hum Reprod 2004 . 15284211
Juul S Karmaus W Olsen J Regional differences in waiting time to pregnancy: pregnancy-based surveys from Denmark, France, Germany, Italy and Sweden. The European Infertility and Subfecundity Study Group Hum Reprod 1999 14 1250 1254 10325272 10.1093/humrep/14.5.1250
Richthoff J Rylander L Jonsson BA Akesson H Hagmar L Nilsson-Ehle P Stridsberg M Giwercman A Serum levels of 2,2',4,4',5,5'-hexachlorobiphenyl (CB-153) in relation to markers of reproductive function in young males from the general Swedish population Environ Health Perspect 2003 111 409 413 12676591
Association WH WHO Laboratory Manual for the Examination of Human Semen and Sperm-Cervical Mucus interaction 1999 Cambridge New York., Cambridge University Press
Toft G Rignell-Hydbom A E T M S Giwercman A Quality control workshops in standardization of sperm concentration and motility assessment in multicentre studies. Int J Androl 2005 15910539
Maldonado G Greenland S Simulation study of confounder-selection strategies Am J Epidemiol 1993 138 923 936 8256780
Inc. SASI SAS System for linear models 1991 Third Cary NC, SAS Institute Inc. 137 191
Rose G Sick individuals and sick populations Int J Epidemiol 1985 14 32 38 3872850
Kelce WR Gray LE Wilson EM Antiandrogens as environmental endocrine disruptors Reprod Fertil Dev 1998 10 105 111 9727600 10.1071/R98051
Safe S Krishnan V Chlorinated hydrocarbons: estrogens and antiestrogens Toxicol Lett 1995 82-83:731-6. 731 736 8597135 10.1016/0378-4274(95)03591-5
Safe S Krishnan V Cellular and molecular biology of aryl hydrocarbon (Ah) receptor-mediated gene expression Arch Toxicol Suppl 1995 17:99-115. 99 115 7786196
Toppari J Environmental endocrine disrupters and disorders of sexual differentiation Semin Reprod Med 2002 20 305 312 12428210 10.1055/s-2002-35377
Fielden MR Halgren RG Tashiro CH Yeo BR Chittim B Chou K Zacharewski TR Effects of gestational and lactational exposure to Aroclor 1242 on sperm quality and in vitro fertility in early adult and middle-aged mice Reprod Toxicol 2001 15 281 292 11390173 10.1016/S0890-6238(01)00129-0
Sager DB Shih-Schroeder W Girard D Effect of early postnatal exposure to polychlorinated biphenyls (PCBs) on fertility in male rats Bull Environ Contam Toxicol 1987 38 946 953 3107640 10.1007/BF01609078
Gray LE Ostby J Furr J Wolf CJ Lambright C Parks L Veeramachaneni DN Wilson V Price M Hotchkiss A Orlando E Guillette L Effects of environmental antiandrogens on reproductive development in experimental animals Hum Reprod Update 2001 7 248 264 11392371 10.1093/humupd/7.3.248
Cocco P Fadda D Ibba A Melis M Tocco MG Atzeri S Avataneo G Meloni M Monni F Flore C Reproductive outcomes in DDT applicators Environ Res 2005 98 120 126 15721892 10.1016/j.envres.2004.09.007
Hauser R Williams P Altshul L Calafat AM Evidence of interaction between polychlorinated biphenyls and phthalates in relation to human sperm motility Environ Health Perspect 2005 113 425 430 15811833
Jonsson B Richthoff J Rylander L Giwercman A Hagmar L Urinary phthalate metabolites in young Swedish males and biomarkers of reproductive function Epidemiology 2005 15951666
Axmon A Rylander L Stromberg U Hagmar L Female fertility in relation to the consumption of fish contaminated with persistent organochlorine compounds Scand J Work Environ Health 2002 Apr;28 (2):124 -32 2002 28 124 132
Mendola P Buck GM Vena JE Zielezny M Sever LE Consumption of PCB-contaminated sport fish and risk of spontaneous fetal death Environ Health Perspect 1995 103 498 502 7656880
Aitken RJ Baker MA Oxidative stress and male reproductive biology Reprod Fertil Dev 2004 16 581 588 15367373 10.1071/RD03089
Beckett GJ Arthur JR Selenium and endocrine systems J Endocrinol 2005 184 455 465 15749805 10.1677/joe.1.05971
Sallmen M Weinberg CR Baird DD Lindbohm ML Wilcox AJ Has Human Fertility Declined Over Time?: Why We May Never Know Epidemiology 2005 16 494 499 15951667 10.1097/01.ede.0000165391.65690.e1
Basso O Juul S Olsen J Time to pregnancy as a correlate of fecundity: differential persistence in trying to become pregnant as a source of bias Int J Epidemiol 2000 29 856 861 11034969 10.1093/ije/29.5.856
Carlsen E Petersen JH Andersson AM Skakkebaek NE Effects of ejaculatory frequency and season on variations in semen quality Fertil Steril 2004 82 358 366 15302284 10.1016/j.fertnstert.2004.01.039
Auger J Eustache F Ducot B Blandin T Daudin M Diaz I Matribi SE Gony B Keskes L Kolbezen M Lamarte A Lornage J Nomal N Pitaval G Simon O Virant-Klun I Spira A Jouannet P Intra- and inter-individual variability in human sperm concentration, motility and vitality assessment during a workshop involving ten laboratories Hum Reprod 2000 15 2360 2368 11056133 10.1093/humrep/15.11.2360
|
16280075
|
PMC1308837
|
CC BY
|
2021-01-04 16:36:33
|
no
|
Environ Health. 2005 Nov 9; 4:26
|
utf-8
|
Environ Health
| 2,005 |
10.1186/1476-069X-4-26
|
oa_comm
|
==== Front
Environ HealthEnvironmental Health1476-069XBioMed Central London 1476-069X-4-271628394110.1186/1476-069X-4-27ResearchInter-population variations in concentrations, determinants of and correlations between 2,2',4,4',5,5'-hexachlorobiphenyl (CB-153) and 1,1-dichloro-2,2-bis (p-chlorophenyl)-ethylene (p,p'-DDE): a cross-sectional study of 3161 men and women from Inuit and European populations Jönsson Bo AG [email protected] Lars [email protected] Christian [email protected] Anna [email protected] Aleksander [email protected] Gunnar [email protected] Henning S [email protected] Jan K [email protected]óralczyk Katarzyna [email protected] Valentyna [email protected]ò Marcello [email protected] Davide [email protected]örgensen Eva C [email protected] Gian Carlo [email protected] Jens Peter [email protected] Lars [email protected] [email protected] Department of Occupational and Environmental Medicine, Lund University Hospital, SE-221 85 Lund, Sweden2 Scanian Fertility Centre, Malmö University Hospital, Malmö, SE-205 02 Sweden3 Department of Occupational Medicine, Aarhus University Hospital, Nørrebrogade 44, build. 2C, DK-8000 Aarhus C, Denmark4 Centre for Arctic Environmental Medicine, Postbox 570DK-3900 Nuuk, Greenland, Denmark5 Department of Environmental Toxicology, National Institute of Hygiene, Warsaw, Chocimska 24, P-00-791 Warsaw, Poland6 Laboratory of Human Reproduction, Kharkiv State Medical University, Klochkovskaya Street 156-A, room 14, 61145 Kharkiv, Ukraine7 Section of Toxicology and Biomedical Sciences, BIOTEC-MED, ENEA CR Casaccia, Via Anguillarese 301, 00060 Rome, Italy8 Istituto di Biologia e Genetica, Università Politecnica delle Marche, Via Brecce Bianche, I-60131 Ancona, Italy9 Institute of Public Health, Department of Environmental and Occupational Medicine, University of Aarhus, University of Aarhus, Vennelyst Boulevard 6, DK-8000 Aarhus C, Denmark10 Laboratorio di Genetica, Dip. di Science Agrarie, Università di Modena e Reggio Emilia, Viele Kennedy 17 – Reggio Emilia I-41100 Modena, Italy2005 11 11 2005 4 27 27 26 7 2005 11 11 2005 Copyright © 2005 Jönsson et al; licensee BioMed Central Ltd.2005Jönsson et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The study is part of a collaborative project (Inuendo), aiming to assess the impact of dietary persistent organochlorine pollutants (POPs) on human fertility. The aims with the present study are to analyze inter-population variations in serum concentrations of 2,2',4,4',5,5'-hexachlorobiphenyl (CB-153) and 1,1-dichloro-2,2-bis (p-chlorophenyl)-ethylene (p,p'-DDE), to assess inter-population variations in biomarker correlations, and to evaluate the relative impact of different determinants for the inter-individual variations in POP-biomarkers.
Method
In study populations of 3161 adults, comprising Greenlandic Inuits, Swedish fishermen and their wives, and inhabitants from Warsaw, Poland and Kharkiv, Ukraine, serum concentrations of CB-153 and p,p'-DDE, were analysed by gas chromatography-mass spectrometry.
Results
The median serum concentrations of CB-153 were for male and female Inuits 200 and 110, for Swedish fishermen 190 and their wives 84, for Kharkiv men and women 44 and 27, and for Warsaw men and women 17 and 11 ng/g lipids, respectively. The median serum concentrations of p,p'-DDE were for Kharkiv men and women 930 and 650, for male and female Inuits 560 and 300, for Warsaw men and women 530 and 380, and for Swedish fishermen 240 and their wives 140 ng/g lipids, respectively. The correlation coefficients between CB-153 and p,p'-DDE varied between 0.19 and 0.92, with the highest correlation among Inuits and the lowest among men from Warsaw. Men had averagely higher serum concentrations of CB-153 and p,p'-DDE, and there were positive associations between age and the POP-biomarkers, whereas the associations with BMI and smoking were inconsistent. Dietary seafood was of importance only in the Inuit and Swedish populations.
Conclusion
CB-153 concentrations were much higher in Inuits and Swedish fishermen's populations than in the populations from Eastern Europe, whereas the pattern was different for p,p'-DDE showing highest concentrations in the Kharkiv population. The correlations between the POP-biomarkers varied considerably between the populations, underlining that exposure sources differ and that the choice of representative biomarkers of overall POP exposure has to be based on an analysis of the specific exposure situation for each population. Age and gender were consistent determinants of serum POPs; seafood was of importance only in the Inuit and Swedish populations.
==== Body
Introduction
The use of persistent organochlorine pollutants (POP), such as polychlorinated biphenyls (PCB), polychlorinated dibenzo-p-dioxins, polychlorinated dibenzofurans and pesticides such as dichlorodiphenyl trichloroethane (DDT) has been restricted and bans implemented in many countries during the 1970's and 1980's, but e.g. DDT is still used in some areas for malaria vector control. These ubiquitous lipophilic POPs are highly resistant to both abiotic and biotic degradation, and are transported long distances by air and water streams.
Some POPs can disrupt multiple endocrine pathways and induce a wide range of toxic responses [1]. A variety of studies have demonstrated their estrogenic, anti-estrogenic, and androgen competing properties [2-4]. There is also concern that POPs may adversely affect male fertility [5].
The present study is a part of a collaborative research project funded by EU (Inuendo; ) which aims to identify and characterize the impact of dietary POP on human fertility in an epidemiological setting [6]. The scope is to contribute scientific knowledge as basis for appropriate assessment and management of risks related to exposure to POP. Study populations have been established in three European countries; Sweden, Poland, and Ukraine together with a population of Inuits from Greenland. The choice of study populations was motivated by that in the Arctic, the combination of environmental conditions and biomagnification in the aquatic food chains result in accumulation of POP in local food at concentrations which are often in excess of contaminants in the mid-latitudes where these contaminants originate [7]. Also Swedish fishermen's families at the east coast at the Baltic Sea, with a high consumption of herring and salmon, constitute a highly exposed group [8,9]. However, in both these populations there are large inter-individual exposure contrasts. Both among Inuits and Swedish fishermen the serum concentrations of e.g. 2,2',4,4',5,5'-hexachlorobiphenyl (CB-153) and 1,1-dichloro-2,2-bis (p-chlorophenyl)-ethylene (p,p'-DDE) were in the same order of magnitude [7-9]. Much less is known about exposure levels for the general populations in Central and Eastern Europe [10]. However, previous data support that the p,p'-DDE concentrations was one order of magnitude higher than the CB-153 concentrations [11].
We have chosen to use CB-153 as a biomarker for POP exposure, because it correlated very well (r = 0.9) with both total PCB concentration in plasma and serum from Swedish subjects and Inuits from Greenland [12-15]. Another relevant biomarker is the anti-androgenic compound p,p'-DDE, the major metabolite of the insecticide DDT with a long biological half-live. Both compounds occur in relatively high concentrations in serum and they are feasible biomarkers in that sense that a large number of analyses can be performed with good precision and accuracy and to a reasonable cost.
The aims with this publication are to a) analyze inter-population variations in CB-153 and p,p'-DDE exposure concentrations and b) assess inter-population variations in biomarker correlations, c) and finally to assess the relative impact of different determinants (gender, age, residence area, breast feeding, smoking, body mass index [BMI], gestational length, and dietary intake of seafood) for the inter-individual variations in serum concentrations of CB-153 and p,p'-DDE.
Subjects and methods
Subjects and data collection
The aim was to recruit 600 pregnant women and their male spouses in Greenland, Warsaw, and Kharkiv, respectively, and 200 Swedish fishermen and 600 Swedish fishermen's wives. A general criterion for eligibility was that the participants had to be born in the country of study and to be at least 18 years of age.
In Greenland pregnant Inuit women from 15 municipalities and 4 settlements from all regions within the country were contacted and informed about the project. In total 893 pregnant women were listed at the local midwives, but 235 did not fulfill the inclusion criteria, 32 could not be contacted and 38 did not want to participate. Blood samples were drawn from the pregnant women and their spouses (all were Inuits), and interviews were made with both partners from June 2002 to December 2003. Of the 598 women included in the time to pregnancy (TTP) study (the results will be published elsewhere), 62 were living alone, 36 men could not be contacted, 34 men did not want to be interviewed, and 5 men did not want to give a blood sample. Moreover, for unknown reasons, blood samples were lost for 16 women and 22 men. Thus, in total, blood samples and interview data were available for 439 men (82 % participation rate) and 572 women (87 % participation rate) (Tables 1 and 2).
Table 1 Serum concentrations of CB-153 and p,p'-DDE (ng/g lipid) and potential confounders among the men.a
Inuit men
(n = 439) Swedish fishermen
(n = 189) Warsaw men
(n = 257) Kharkiv men
(n = 287)
Mb 5–95 % M 5–95 % M 5–95 % M 5–95 %
CB-153 200 50–920 190 62–640 17 <LODc-38 44 5.5–130
p,p'-DDE 560 90–2200 240 80–890 530 220–1000 930 390–2400
Age (years) 30 21–44 48 32–62 30 25–37 25 20–39
BMI (kg/m2) 25.6 20.4–33.1 26.4 22.0–32.6 25.4 20.4–31.7 23.9 19.7–29.5
Smoking
Neverd 16 38 52 25
Everd 18 62 48 75
Cig/day 8.0 0–20 11 0–25 0 0–20 7.5 0–20
Seafood (meals/week) 1.5 0.1–5.0 n.ae 1.0 0–3.0 4.0 2.0–5.0
a) Except area of residence
b) Median
c) Below detection level
d) %
e) n.a. = not analyzed because data was not available
Table 2 Serum concentrations of CB-153 and p,p'-DDE (ng/g lipid) and potential confounders among the women.a
Inuit women
(n = 572) Swedish women
(n = 544) Warsaw women
(n = 261) Kharkiv women
(n = 612)
Mb 5–95 % M 5–95 % M 5–95 % M 5–95 %
CB-153 110 21–530 84 30–220 11 <LODc-27 27 8.1–69
p,p'-DDE 300 49–1300 140 50–530 380 140–880 650 270–1700
Age (years) 27 19–38 50 37–57 29 25–34 24 18–34
BMI (kg/m2) 23.8 18.9–32.8 24.8 20.3–32.9 21 18.1–27.3 24 17.6–27.7
Smoking n.a.e
Neverd 16 81 82
Everd 84 19 18
Cig/day 6 0–15 0 0–10 0 0–10
Seafood (meals/week) 1.2 0–5.0 0.4 0–1.3 1.0 0–3.0 0 0–3.0
Gestational length (weeks) 24 9–38 n.a. 33 9–38 22 7–41
Primiparad 32 n.a. 93 80
Total lactation time (months) n.a. 12 0–40 n.a. n.a.
a) Except area of residence
b) Median
c) Below detection level
d) %
e) n.a. = not analyzed because data was not available
Altogether 690 pregnant women and their spouses who either visited the obstetric out-patient clinic of the Gynecological and Obstetric Hospital of the Warsaw Medical University, Poland, or physicians at a collaborating hospital in the same city were informed about the project and asked to participate. Of those, 261 (38 % participation rate) women and 257 (37 % participation rate) men donated a blood sample and were interviewed from September 2002 to March 2003 (Tables 1 and 2). The population was urban.
Altogether 2478 pregnant women and their spouses who visited one of eight antenatal clinics or three maternity hospitals in Kharkiv, Ukraine, were informed about the project and asked to participate. Of those, 612 (25 % participation rate) women and 287 (12 % participation rate) men donated a blood sample and were interviewed from April 2003 to December 2004. The population had a mixed urban and rural background.
In contrast to the other three participating populations, the recruitment process in Sweden was not focused on ongoing pregnancies. Instead blood samples were drawn from 189 professional fishermen from the Swedish east and west coasts cohorts that participated in a semen study between March 2001 and November 2001, and March 2002 and September 2002 (Tables 1 and 2). Initially 2614 Swedish fishermen had been informed about the semen study and 266 (10 % participation rate) gave their written informed consent to participate, but 71 subjects had to be excluded during the sampling period due to logistical reasons, changes of mind, sickness or recent vasectomy during the field study period [16]. Altogether 1439 fishermen's wives, born 1945 or later, were contacted for a TTP study. Blood samples were drawn from, and interviews were performed on 544 (38 % participation rate) women between December 2002 and March 2004. Eighty-seven spouses to the fishermen that participated in the semen study were eligible and 68 of them participated.
The study was approved by the local ethical committees representing all participating populations and all subjects signed an informed consent.
Interview data
Information on lifestyle, medical and reproductive history was collected through interviews. In the present study we used information on age, gestational length, height and weight (for calculation of BMI), smoking habits, and intake of seafood. For the Inuits seafood comprised sea mammals as well as fish, whereas for the other populations it only refers to fish. For the fishermen's wives from Sweden information on life-long total length of breast-feeding was obtained and for the other women whether it was the first time they were given birth (primipara). For the populations from Sweden and Greenland we also used information regarding area of living. Information on current smoking was not obtained for the women from Sweden and information on current intake of seafood was not obtained for the Swedish fishermen.
Collection of blood samples
Blood samples were drawn from a cubital vein into 10 ml vacuum tubes for serum collection without additives (Becton Dickinson, Maylan, France). After cooling to room temperature the tubes were centrifuged at 4000 g for 15 min. Serum was transferred with ethanol rinsed Pasteur pipettes to ethanol rinsed brown glass bottles (Termometerfabriken, Gothenburgh, Sweden). A piece of aluminum foil was placed on top of the bottles which were then sealed. Sera were stored at -20°C until shipment, but it was accepted to keep it in refrigerator for up to four days.
Determination of CB-153 and p,p'-DDE in serum
The sera were transported on dry ice to the Department of Occupational and Environmental Medicine in Lund, Sweden, where all analysis of CB-153 and p,p'-DDE were performed. The CB-153 and p,p'-DDE were extracted from serum by solid phase extraction using on-column degradation of the lipids and analysis by gas chromatography mass spectrometry as previously described [16]. The relative standard deviations, calculated from samples analyzed in duplicate at different days, was 18% at 0.1 ng/mL (n = 990), 10 % at 0.5 ng/mL (n = 990) and 10 % at 2 ng/mL (n = 990) for CB-153 and 11 % at 1 ng/mL (n = 1058), 8 % at 3 ng/mL (n = 1058) and 7 % at 8 ng/mL (n = 1058) for p,p'-DDE. The detection limits were 0.05 ng/mL for CB-153 and 0.1 ng/mL for p,p'-DDE. The analyses of CB-153 and p,p'-DDE were part of the Round Robin inter-comparison program (Prof. Hans Drexler, Institute and Out-Patient Clinic for Occupational, Social and Environmental Medicine, University of Erlangen-Nuremberg, Germany) with analysis results within the tolerance limits. An internal control at 0.4 ng/mL for CB-153 and 0.6 ng/mL for p,p'-DDE was included in all analyzed series. The samples were analyzed in duplicate and the samples were re-analyzed a third time if the difference between the two determinations were higher than 30 %. However, at concentrations below 0.2 ng/mL an absolute deviation of 0.1 ng/mL between the duplicate samples was allowed. After the third analysis the two determinations fulfilling the above criteria was chosen. If none of the determinations fulfilled the criteria the sample were analyzed a fourth time.
Determination of lipids by enzymatic methods
Serum concentrations of triglycerides and cholesterol were determined by enzymatic methods using reagents from Roche Diagnostics (Mannheim, Germany). The inter-assay CVs for cholesterol and triglyceride determinations were 1.5–2.0 %. The average molecular weights of triglycerides were assumed to be 807. For cholesterol we used an average molecular weight of 571, assuming that the proportion of free and esterified cholesterol in plasma was 1:2. Based on a paper by Rylander et al., the total lipid concentration in serum (g/L) was calculated by the following equations [17]:
Men: Total = 0.96 + 1.28*(triglycerides + cholesterol)
Women: Total = 1.13 + 1.31*(triglycerides + cholesterol).
Statistical analysis
Separate analyses for each population were performed for men and women. If the concentration was less than the limit of detection, the concentration was set to half of the detection limit based on the fresh weight concentrations. For CB-153 there were 122 (men: 37; women: 85) and for p,p'-DDE there were 13 (men: 3; women: 10) values below the limits of detection. The impact of potential determinants for inter-individual variations in CB-153 and p,p'-DDE serum concentrations were assessed by linear regression models. We also evaluated whether log transformations of the POP variables better fulfilled the model assumptions, which were checked by analysis of residuals according to the methods suggested by Altman [18]. The determinants evaluated were age (continuous), BMI (continuous), smoking (categorical: ever/never; continuous: cigarettes/day), and number of meals of seafood per week (not for Swedish fishermen). The impact of area of residence was assessed for the Inuits (three categories: Living in the capital Nuuk [presumed low exposure], living on the remaining west coast [presumed medium exposure] and living in the north or east coast regions [presumed high exposure]) and the Swedish study population (two categories: west [presumed low exposure] and east [presumed high exposure] coasts). In addition, we also evaluated the impact of gestational length at blood sampling and parity (primipara: yes/no) for women from Greenland, Warsaw and Kharkiv. Among Swedish fishermen's wives the impact of life-long total length of breast-feeding was assessed (four categories: 0–6, >6–12, >12–24 and >24 months).
The objective of the multivariate model building was to evaluate the impact of each of the determinants. In the first step we evaluated the associations for one determinant at a time with the CB-153 and p,p'-DDE concentrations in serum, respectively. In the second step we kept the determinant with the lowest p-value (if ≤ 0.15) and tentatively included all other determinants, one at a time. In the third step we kept the two determinants with the lowest p-values (if both ≤ 0.15) and tentatively included the remaining determinants, one at a time. This procedure was continued as long as the p-values for the all included determinants were ≤0.15. For the populations from Greenland and Sweden (only among the women) we also evaluated whether area of living modified the effect of seafood by including an interaction term in the models. The explained variances (adjusted r2 obtained from SPSS™), after successive addition of determinants to the final multivariate model, are given.
The inter-population variations in CB-153 and p,p'-DDE serum concentrations were also assessed by linear regression models. In these models age was considered as a potential confounder. The age distribution among the women from Sweden differed so much from the other age distributions (with almost no overlap) and we did therefore not make age-adjusted comparisons for the women from Sweden.
The correlations between serum concentrations of CB-153 and p,p'-DDE for the different populations were evaluated by Spearman's correlation coefficients.
Results
Inter-population variations in serum concentrations of CB-153 and p,p-DDE
In total, the material included lipid adjusted serum concentrations of CB-153 and p,p'-DDE for 1172 men and 1989 women. In Tables 1 and 2, the concentrations and the distributions of CB-153 and p,p'-DDE in the different populations are given. In Table 3 the serum concentrations of the POP markers are given by area of residence within Greenland and Sweden.
Table 3 Serum concentrations of CB-153 and p,p'-DDE (ng/g lipid) with respect to area of residence. Data for men and women from Greenland and Sweden, respectively.
Men Women
n Ma 5–95 % n M 5–95 %
CB-153
Greenland
Nuuk 122 130 38–680 159 79 11–328
Westb 286 220 55–810 377 120 23–490
East/Northc 29 600 110–1900 33 490 77–1400
Sweden
West coast 97 170 55–370 360 79 32–160
East coast 92 210 63–800 184 97 27–260
p,p-DDE
Greenland
Nuuk 122 340 47–1600 159 240 15–780
Westb 286 610 150–2200 377 310 59–1100
East/Northc 29 1900 310–5000 33 1000 230–2400
Sweden
West coast 97 200 62–620 97 200 62–620
East coast 92 290 110–1200 92 290 110–1200
a) median
b) Aasiaat, Ilulissat, Kangaatsiaq, Maniitsoq, Nanortalik, Narsaq, Paamiut, Qaqortoq, Qasigiannguit, Qeqertarsuaq, Saattut, Sisimiut, Ukkusissat and Uummannaq
c) Qaanaaq, Tasiilaq, Kulusuk and Kuummiut
For the men the mean serum concentrations of CB-153 was in the following order: Inuits (highest), Swedish fishermen, Warsaw and Kharkiv (lowest) (Table 1). The crude mean difference between Inuits and Swedish fishermen was 78 ng/g lipid (95 % CI 19, 137, p < 0.001), and this difference became even more evident after age adjustment (mean difference 297 ng/g lipid; 95 % CI 220, 375, p < 0.001), indicating that the Inuits had reached high serum concentrations of CB-153 already in their twenties. The fishermen from Sweden had higher concentrations as compared with the men from Kharkiv (age adjusted mean difference: 43, 95 % CI 3.5, 82), which in turn had higher concentrations as compared with the men from Warsaw (age adjusted mean difference 43 ng/g lipid, 95 % CI 33, 53).
The pattern among the males with respect to population was quite different for p,p'-DDE (Table 1). The men from Kharkiv had significantly higher concentrations than the Inuits (age adjusted mean difference 457 ng/g lipid, 95 % CI 309, 605), which in turn had significantly higher concentrations than the men from Warsaw (age adjusted mean difference 213 ng/g lipid, 95 % CI 96, 329), and finally the Swedish fishermen had significantly lower concentrations than the men from Warsaw (age adjusted mean difference 426 ng/g lipid, 95 % CI 339, 514).
The Inuit women had significantly higher serum concentrations of CB-153 as compared with the women from Kharkiv (age-adjusted mean difference 133 ng/g lipid, 95 % CI 116, 150), whom in turn had higher concentrations than the women from Warsaw (age-adjusted mean difference 24 ng/g lipid, 95 % CI 20, 28).
The women from Kharkiv had significantly higher age-adjusted serum concentrations of p,p'-DDE as compared with the Inuit women and the women from Warsaw (all p-values<0.001). There was, however, no age-adjusted difference between the Inuit women and the women from Warsaw (p = 0.17).
Inter-population variations in biomarker correlations
The correlation coefficients (rs) between serum concentrations of CB-153 and p,p'-DDE in the different populations varied between 0.19 and 0.92, with the highest correlation among Inuits and the lowest correlation among men from Poland (Table 4).
Table 4 Correlations between serum concentrations of CB-153 and p,p'-DDE.a
Inuits Swedish fishermen's populations Warsaw Kharkiv Total
Men, rsb 0.93 0.79 0.19 0.42 0.17
Women, rsb 0.92 0.76 0.54 0.49 0.03
a) For men and women for each of the populations, and for the total study group
b) Spearman's correlation coefficient
Impact of determinants for inter-individual variations in serum POP concentrations
There was a positive association between age and serum concentrations of CB-153 in all male populations, although the impact of an increase in age of one year meant more in Inuits (β = 16 ng/g lipid) and Swedish fishermen (β = 9.0) as compared with men from Warsaw (β = 1.2) and Kharkiv (β = 2.0) (Tables 5, 6, 7, 8). The quantitative effect of age can be exemplified by that the serum concentrations of CB-153 increased in the Inuits by 16 ng/g lipids for each extra year of age, whereas the similar age-related increase was only 1.2 ng/g lipids among the men from Warsaw. Age was also positively associated (p < 0.001) with p,p'-DDE serum concentrations in the Inuit and Swedish populations (Tables 5 and 6). In concordance with the results for men, also among the women age was positively associated with POP concentrations in all populations (all p-values = 0.01, Tables 5, 6, 7, 8).
Table 5 Linear multiple regressions for determinants for serum concentrations of CB-153 (ng/g lipid) and p,p'-DDE (ng/g lipid) among Inuit men and women.a
Inuit men Inuit women
CB-153 p,p'-DDE CB-153 p,p'-DDE
Age, years 16 (12, 21) 31 (20, 42) 7.7 (5.2, 10) 15 (9.0, 21)
BMI, kg/m2 -b - -2.9 (-6.1, 0.2) -b
Current smoking
No - - Ref Ref
Yes - - 36 (-1.5, 75) 101 (13, 189)
Cig/day - - - -
Seafood, meals/week 53 (33, 73) 130 (80, 179) 18 (8.8, 28) 58 (36, 80)
Area
Nuuk Ref Ref Ref Ref
West 87 (13, 162) 309 (127, 491) 46 (14, 78) 94 (20, 168)
East/North 597 (454, 740) 1271 (922, 1620) 399 (335, 463) 773 (626, 921)
Primipara
No Ref Ref
Yes 26 (-7.5, 60) 60 (-18, 138)
Gestational length (weeks) - -
Explained variancec(%) 12 (area)
24 (+age)
27 (+seafood) 12 (area)
18 (+age)
22 (+seafood) 21 (age)
26 (+area)
28 (+cig/day) 16 (area)
21 (seafood)
24 (+age)
25 (+smoking)
25 (+parity)
a) The β-coefficients (with 95 % CI within brackets) are shown.
b) The variable did not fulfil the criteria to be included in the final model.
c) Cumulative explained variance after successive addition of determinant variables to the final multivariate model. Thus, the total explained variance is given on the lowest line.
Table 6 Linear multiple regressions for determinants for serum concentrations of CB-153 (ng/g lipid) and p,p'-DDE (ng/g lipid) among Swedish fishermen and their wives.a.
Swedish fishermen Fishermens' wives
CB-153 p,p'-DDE CB-153 p,p'-DDE
East coast West coast East coast West coast
Age, years 9.0
(6.5, 12) 10
(6.1, 14) 6.3
(4.7, 8.0) 2.9
(2.2, 3.6) 18
(13, 22) 5.4
(3.7, 7.2)
BMI, kg/m2 -b 13
(0.3, 25) - - 6.2
(-0.7, 13) 3.0
(-0.1, 6.1)
Current smoking n.a.c n.a. n.a. n.a.
No - -
Yes - -
Cig/day -2.9
(-5.1, -0.6) -
Seafood, meals/week n.a. 49
(23, 75) 17
(6.4, 28) 112
(35, 189) -
Area
West coast Ref Ref
East coast 96
(48, 143) 160
(80, 241)
Lactation (months) - -
0–6 Ref Ref
>6–12 3.7
(-23, 30) 1.7
(-11, 14)
>12–24 10
(-16, 37) -16
(-29, -5.0)
>24 -35
(-64, -5) -33
(-46, -19)
Explained varianced(%) 21 (age)
26 (+area)
28 (+cig/day) 12 (age)
20 (+area)
21 (+BMI) 28 (age)
33 (+seafood)
36 (+lactation) 20 (age)
25 (+lactation)
27 (+seafood) 21 (age)
26 (+area)
28 (+cig/day) 16 (area)
21 (seafood)
24 (+age)
25 (+smoking)
25 (+parity)
a) The β-coefficients (with 95 % CI within brackets) are shown.
b) The variable did not fulfil the criteria to be included in the final model.
c) n.a = not analyzed because data was not available.
d) Cumulative explained variance after successive addition of determinant variables to the final multivariate model. Thus, the total explained variance is given on the lowest line.
Table 7 Linear multiple regressions for determinants for serum concentrationsof CB-153 (ng/g lipid) and p,p'-DDE (ng/g lipid) among Warsaw men and women.a
Warsaw men Warsaw women
CB-153 p,p'-DDE CB-153 p,p'-DDE
Age, years 1.2 (0.8, 1.6) - 1.3 (0.9, 1.6) 21 (9.6, 32)
BMI, kg/m2 -0.6 (-1.1, -0.1) -10 (-21, 0.8) - -
Current smoking -a -
No - - - Ref
Yes - -90 (-176,-3.9)
Cig/day - - -0.2 (-0.4, 0.05) -
Seafood, meals/week - - 1.1 (0.1, 2.1) -
Primipara -
No Ref
Yes 7.4 (2.8, 12)
Gestational length (weeks) - -
Explained variancec(%) 12 (area)
24 (+age)
27 (+seafood) 12 (area)
18 (+age)
22 (+seafood) 14 (age)
19 (+parity)
19 (+seafood)
19 (+cig/day) 6 (age)
7 (+smoking)
a) The β-coefficients (with 95 % CI within brackets) are shown.
b) The variable did not fulfil the criteria to be included in the final model.
c) Cumulative explained variance after successive addition of determinant variables to the final multivariate model. Thus, the total explained variance is given on the lowest line.
Table 8 Linear multiple regressions for determinants for serum concentrations of CB-153 (ng/g lipid) and p,p'-DDE (ng/g lipid) among Kharkiv men and women. a.
Kharkiv men Kharkiv women
CB-153 p,p'-DDE CB-153 p,p'-DDE
Age, years 2.0 (0.4, 3.5) - 1.4 (0.8, 2.0) 19 (7.6, 31)
BMI, kg/m2 -b - -0.8 (-1.6, -0.1) -
Current smoking
No - Ref - -
Yes - -200 (-458, 57) - -
Cig/day - - - -
Seafood, meals/week - - - -
Primipara
No Ref Ref
Yes 5.5 (-1.5, 12) 135 (-2.7, 273)
Gestational length (weeks) -0.3 (-0.5, -0.1) -7.9 (-12, -4.0)
Explained variancec(%) 2 (age) 1 (cig/day) 3 (age)
4 (+gestational length)
5 (+BMI)
5 (+parity) 2 (gestational length)
4 (+age)
4 (+parity)
a) The β-coefficients (with 95 % CI within brackets) are shown.
b) The variable did not fulfil the criteria to be included in the final model.
c) Cumulative explained variance after successive addition of determinant variables to the final multivariate model. Thus, the total explained variance is given on the lowest line.
Regarding BMI and smoking the pattern was very inconsistent, e.g. the effect of BMI on serum concentrations of p,p'-DDE went in different directions among the Swedish fishermen and the men from Warsaw (Tables 6 and 7). Also among the women the patterns for BMI and smoking with respect to POP concentrations were very inconsistent (Tables 5, 6, 7, 8).
In both Inuit men and women and among the Swedish fishermen the area of residence was of highly significant importance (all p-values<0.001) for the POP concentrations (Tables 5 and 6).
Among Inuit men the number of seafood meals per week was positively associated with the serum concentrations of CB-153 (p < 0.001) as well as with the serum concentrations of p,p'-DDE (p < 0.001) (Table 5). Similar significant associations (p < 0.001) were seen among the Inuit women, and also for serum concentrations of CB-153 among the women from Warsaw (p = 0.03) (Table 7). Among the Swedish fishermen's wives there was a significant interaction (p < 0.001) between intake of seafood and area of residence (Table 6). The intake of seafood had a major impact on fishermen's wives from the Swedish east coast, whereas the association was less obvious among the west coast women. There was no such interaction between intake of seafood and area of residence among the Inuit women.
The women who were giving birth to their first child had in general higher concentrations of POP concentrations, although most p-values were above 0.05 (Tables 5, 7, 8). In the Swedish population, the fishermen's wives with the longest total life-time length of breast-feeding had lower concentrations of POP, with the exception of p,p'-DDE concentrations among the east coast women (Table 6). A significant negative association between gestational length at blood sampling and POP in serum was observed only for women from Kharkiv, corresponding to a weakly decrease of CB-153 with 0.3 ng/g and of p,p'-DDE with 7.9 ng/g (Table 8).
The explained variances in the final regression models among the men from the different populations varied between 1 and 28 %, and for the different female populations it varied between 4 and 36 % (Tables 5, 6, 7, 8). By log transformation of the POP variables the model assumptions were somewhat better fulfilled in the Swedish population. However, in the three other populations no such improvements were achieved by log transformation of the POP variables, and we do therefore present the results for the untransformed variables for all four populations.
Discussion
As could be expected serum concentrations of CB-153 were much higher in the Inuit and Swedish fishermen's populations as compared with the populations from Eastern Europe, whereas the exposure pattern was different for p,p'-DDE showing highest concentrations in the Kharkiv population and relatively low concentrations in the Swedish fishermen's populations. It has to be borne in mind that the participants from Sweden were not recruited from the general Swedish population but from a fishermen's population with averagely higher POP exposures. This was done in order to ensure a sufficient exposure contrast within the total study population. Moreover, the age distribution in the Swedish fishermen's populations differed a lot from the age distributions among the populations from the other countries. Thus, direct comparisons between the Swedish and the other populations must therefore be interpreted with caution.
The exposure pattern in the Kharkiv population with relatively high p,p'-DDE concentrations and relatively low CB-153 concentrations are in accordance with a previous analysis of maternal milk from Ukraine, donated 1993–1994, in which the median p,p'-DDE concentration (2457 ng/g lipid) was one order of magnitude higher than the median CB-153 concentration (149 ng/g lipid) [11]. The most reasonable explanation for this is that the use of the insecticide DDT has ceased relatively late in the Ukraine.
A noteworthy result of the present study was that the degree of correlation between the two POP biomarkers, CB-153 and p,p'-DDE, varied considerably between the different populations. The extremely high correlation coefficients (rs 0.92 and 0.93, respectively; Table 4) for the Inuits and the high correlation coefficients for the Swedish fishermen's populations (rs 0.79 and 0.76, respectively) support a common exposure source for both contaminants. This was not that obvious for the populations from Warsaw and Kharkiv, for which there were no previous data concerning the inter-correlation between CB-153 and p,p'-DDE. These large differences in correlation underline that the choice of representative biomarkers of overall POP exposure has to be based on an analysis of the specific exposure situation for each population. It might seem paradoxical that the correlation coefficients between the two exposure biomarkers became considerably lower when pooling the four populations, but the reason is that the exposure ranges did vary quite substantially between the different populations. Undoubtedly, it would have been valuable to have a more extensive exposure profile of the POPs than the two biomarkers analysed in this project, and we plan to perform such analyses on a stratified sub-sample from the study populations. On the other hand, it should be kept in mind that for practical and economical reasons it would have been impossible to analyse more than 3,000 samples, as in the present project, for a more detailed POP exposure profile. When performing epidemiological studies on reproductive outcomes, large enough sample sizes are needed to ensure a reasonable statistical power. Thus, a trade off is needed between costs for obtaining large enough number of samples and costs for detailed analyses of each sample, in order to obtain the most efficient study design.
The analysis of the relative impact of different determinants for variation in serum concentrations of the POP biomarkers showed a consistent association with gender. For the four geographical regions the median CB-153 concentrations were 55–126 % higher among the men. The corresponding figures for p,p'-DDE were 39–87 %. The results are in concordance with previous studies of Inuits [19], frequent consumers of Great Lakes sport fish [20], and Swedish fishermen's families [7,8]. The main reason for the observed gender difference is probably excretion of POP by breast feeding.
Increasing age was also a most consistent determinant of higher serum concentrations of CB-153 and p,p'-DDE in both men and women. This is in line with earlier findings [20-23], and could be caused by both an age-dependent bioaccumulation of the persistent and lipophilic compounds and higher exposure concentrations in the past detectable in the more elderly; a birth cohort effect [24]. Previous Swedish data on women support that the birth cohort effect is of great importance [22]. This is due to that the highest POP contaminant concentrations in different food items were found in the 1970's and the concentrations have continuously decreased since then. The effect of this is that for a given calendar year of birth, CB-153 in plasma decreased with age, from the time point when steady state of body burden was reached for this compound with a biological half-live of many years, and onwards. Thus, in a longitudinal study design, the plasma concentration of CB-153 will decrease with age, from time of steady state, while applying a cross-sectional approach, older subjects will have higher concentrations than younger ones.
When evaluating the impact of age on serum concentrations of POP in women, one has to keep in mind that age is associated with parity and therefore also with total length of breast-feeding, which works against an observed increase in serum concentrations of POP with age. Consequently, there was in the present study an independent negative association between life-long total length of breast-feeding and POP concentrations in serum among the Swedish women. In the multivariate models breast-feeding increased the explained variance for the POP biomarkers among the Swedish women with, as most, 6 %. Negative associations between total length of breast-feeding and PCB concentrations in mother's milk [25-27], as well as blood plasma [22] have previously been shown. About a third of the Inuit women were primipara, and parity did only increase the explained variance in the multivariate models with 1–2 %. The exposure determinant seafood varied between the populations. Among Inuits seafood comprised both sea mammals and fish, whereas for Sweden it comprised locally caught fish from the Baltic Sea (east coast) and the North Sea (west coast), respectively. For both male and female Inuits there were clear positive association between consumption of seafood and the POP markers. Among the fishermen's wives from Sweden the area of living modified the effect of fish consumption on POP concentrations. The impact of seafood (per weekly meal) was greater among women from the Swedish east coast, which is explained by the higher concentration of POP contamination in fatty fish from the Baltic Sea as compared with fish from the North Sea at the west coast [28]. In a previous Swedish study on elderly women from the general population [29] and in US studies [24], fish consumption was associated with exposure to PCBs but not to p,p'-DDE. A positive association between fish consumption and PCB was observed for both genders among high consumers of sport fish from the Great Lakes [20]. The seafood consumed in the Warsaw and Kharkiv populations was mainly fish from freshwater lakes and this did not seem to be a substantial exposure source for CB-153 or p,p'-DDE.
There was in the present study no consistent association between BMI and serum concentrations of CB-153 or p,p'-DDE. The positive association between BMI and p,p'-DDE among the Swedish fishermen was the only significant finding, but the explained variance in the multivariate model did only increase with 1 % by including BMI in the model. In a previous study POPs were negatively associated with BMI among young and middle-aged male Inuits from Greenland [14], but we did not observe such an association. The previous literature gives no clear picture of the association between BMI and body burdens of POP. BMI was a predictor of serum p,p'-DDE in some studies [21,30,31], but not in other [24,32]. Two studies on women from the US found a negative association between BMI and serum concentration of higher chlorinated or total PCB [23,31], whereas no association was found between total PCB concentration and BMI in two other studies [21,24]. A positive association between BMI and PCB was observed for both genders among high consumers of sport fish from the Great Lakes [20]. The association with BMI varied between congeners in a Swedish study of elderly women [29]. For some PCB-congeners (CB-105 and CB-118) and p,p'-DDE positive associations were seen, while negative associations were observed for some other PCB congeners (CB-156 and CB-180), and for e.g. CB-153 no significant association at all was observed. Thus, there may be compound-specific differences in the modulating effects of BMI on serum POP concentrations. Moreover, the inconsistent results with respect to the association between BMI and POP in serum may be explained by the timing of blood sampling in relation to when the period of more substantial dietary POP exposure had taken place [33]. If this exposure had been recent, subjects with high BMI, i.e. with large adipose distribution volumes, should be expected to have negative association between BMI and lipid adjusted POP concentrations in serum. On the other hand, if the blood samples were drawn many years after end of a more substantial exposure than the present one, and the subjects were old enough to have reached a steady state of their body burdens of POPs, considering the long biological half-lives of many of these compounds, you could expect to find a positive association between BMI and POP in serum. This reasoning fits with our finding of a positive association between BMI and p,p'-DDE among the middle-aged and elderly Swedish fishermen, who decades ago had been substantially more exposed to POPs through consumption of fatty fish from the Baltic Sea [28].
Among the Inuit women we observed positive associations between current smoking and serum concentrations of both CB-153 and p,p'-DDE. However, by including smoking in the multivariate model the explained variance did only increase with 1 %. In contrast, a negative association between current smoking and p,p'-DDE was observed among the women from Warsaw, whereas no association between POP markers and smoking was observed among women from Kharkiv. In Inuit men and men from Kharkiv, smoking was not associated with any of the POP biomarkers, whereas there was a negative association between smoking and CB-153 among Swedish fishermen. Thus, is it possible to explain this ambiguous pattern of associations? Our results are in accordance with previous studies of Inuit women from Greenland [14,34], and also of Inuit women from Canada [35]. It is, however, of interest that no such association was found among the Caucasian control women in the latter study. Moreover, in two Swedish studies on elderly women and fishermen's wives, respectively, no association between POP and smoking was seen [22,29]. In contrast to the present results, previous studies on male Inuits from Greenland showed in multivariate analyses positive associations between smoking and PCB in serum [37]. It has been proposed that smoking can affect uptake or metabolism of POPs, e.g. by influencing the CYP-450 enzyme system [14], but this seems to be an unlikely explanation as induced CYP-450 system (by smoking or PCB) would rather result in lower PCB concentrations due to increased formation of hydroxylated metabolites. It has been reported an association between consumption of sea mammal fat and tobacco smoking among Canadian Inuit [15], and a more probable explanation for the presently observed weak association between smoking and POP in Inuit women might be a residual confounding of diet.
The participation rate varied considerably between the different populations (10–87 %). We have, however, no reason to suspect that the subject's decision to participate was affected by any knowledge of their individual POP exposure. It was not possible to adjust for participation rate as this is a population characteristic and not an individual characteristic. Concerning the external validity of the present study, our conclusions are restricted to the age groups studied. Thus, we cannot say anything about serum concentrations of POP biomarkers in children, adolescents or in women past their reproductive age (except for Sweden). Three of the female study populations comprised pregnant women, which could hamper a comparison with non-pregnant women of the same age. However, gestational length affected the POP biomarker concentrations only in the Kharkiv population. Adjusting for the other determinants, each extra week of pregnancy in Kharkiv women was associated with a decrease of the POP biomarkers with 1 %. We think that the participating men from Greenland, Warsaw and Kharkiv were representative for the general population in the same geographical area, whereas the Swedish fishermen were more highly exposed to POPs than the general Swedish population.
We used CB-153 and p,p'-DDE in serum as index biomarkers of exposure to POPs and there are several previous studies supporting this approach [12-15]. It must, however, be taken into consideration that the relative concentrations of non-coplanar and co-planar PCBs can differ between regions depending on varying exposure sources. Studies of Inuit populations from Canada support in general the use of CB-153 as a surrogate marker of exposure to non-dioxin like PCBs present in the Arctic food-chain [15]. However, the ratio between serum concentrations of CB-153 (and other non-coplanar congeners) and co-planar PCBs was higher in Canadian Inuits than in Canadian Caucasians from the Arctic area [37], which calls for caution using CB-153 as a global exposure marker for POP.
POP concentrations in blood have previously been analyzed in several studies both from Greenland and Sweden. Samples drawn in 1994–1997 from pregnant women from the Disco Bay area at the west coast of Greenland showed that both CB-153 and p,p'-DDE were at that time found in about 50 % higher concentrations than in the present study [7,38,39]. Glynn et al. monitored in 1996–1997 CB-153 and p,p'-DDE in serum from elderly women (54–75 years) living on the Swedish east coast or around the Swedish large lakes [29]. The concentrations were about two times that of the Swedish fishermen's wives in the present study. These results are in accordance with other studies also showing a trend of decreasing POP exposure during the last decades [40-47]. Asplund et al. monitored in 1987 CB-153 and p,p'-DDE in plasma from Swedish men (25–56 years) with low to high consumption of fatty fish from the Baltic Sea [48]. The Swedish low consumers in 1987 had POP concentrations twice as high as the Swedish fishermen in the present study. A similar study, with samples collected in 1991, showed among men with low consumption of fatty fish from the Baltic Sea, CB-153 and p,p'-DDE in the same range as those presently reported for Swedish fishermen, while the concentrations in the high consumer group was more than twice as high [49]. Thus, these results indicate that there has been a decrease in POP concentrations in blood in both Greenland and Sweden during the last decade(s). For Warsaw and Ukraine there are no reports of CB-153 and p,p'-DDE in blood that can be used for time trend analyses. However, analysis of maternal milk from Ukraine, donated 1993–1994, showed a median p,p'-DDE concentration of 2457 ng/g lipid [11], which is considerably higher than the median lipid adjusted p,p'-DDE concentration in serum of 650 ng/g lipid in the present study, indicating a rather dramatic decrease in exposure during the last decade.
Conclusion
Serum concentrations of CB-153 were much higher in the Inuit and Swedish fishermen's populations as compared with the populations from Eastern Europe, whereas the exposure pattern was different for p,p'-DDE showing highest concentrations in the Kharkiv population and relatively low concentrations in the Swedish fishermen's populations.
The degree of correlation between the two POP biomarkers varied considerably for the different populations, which underlines that the exposure sources obviously differ between countries and that the choice of representative biomarkers of overall POP exposure has to be based on an analysis of the specific exposure situation for each population.
Age and gender affected the serum POP concentrations in all populations whereas the associations with BMI and smoking were inconsistent. Dietary seafood and area of residence were of importance for the POP biomarker concentrations only in the Inuit and Swedish populations.
List of Abbreviations
BMI body mass index
CB-153 2,2',4,4',5,5'-hexachlorobiphenyl
CV coefficient of variation
p,p'-DDE 1,1-dichloro-2,2-bis (p-chlorophenyl)-ethylene
DDT dichlorodiphenyl trichloroethane
PCBs polychlorinated biphenyls
POPs persistent organochlorine pollutants
TTP time to pregnancy
Competing interests
The author(s) declare that they have no competing interest.
Authors' contributions
JPB, AG and LH initiated the project. JPB was main responsible for raising funding for the project. ARH, GT, HSP, JKL, KG, and VZ have been responsible for collecting all blood samples and for obtaining the interview data. BAGJ and CL were responsible for the chemical analyses of the POP biomarkers. GT had main responsibility for creating the joint database. LR performed the statistical analyses. LH, BAGJ, LR, JPB, GT and ARH helped to draft the manuscript. All authors participated in the design of the study and have read and approved the final manuscript.
Acknowledgements
This study is part of the project "Inuendo – Biopersistent organochlorines in diet and human fertility. Epidemiological studies of time to pregnancy and semen quality in Inuit and European populations", supported by The European Commission to the 5th Framework Programme Quality of Life and Management of Living Resources, Key Action 4 on Environment and Health (Contract no. QLK4-CT-2001-00202, ). The work has also been funded by the Swedish Research Council and the Swedish Research Council for Environment, Agricultural sciences and Spatial Planning. Authors thank Ms Hélène Åkesson, Ms Berit Holmskov, and Ms Christina Held for performing chemical analyses in a skilful way.
==== Refs
Toppari J Larsen JC Christiansen P Giwercman A Grandjean P Guillette LJ JrJegou B Jensen TK Jouannet P Keiding N Leffers H McLachlan JA Meyer O Muller J Rajpert-De Meyts E Scheike T Sharpe R Sumpter J Skakkebaek NE Male reproductive health and environmental xenoestrogens Environ Health Perspect 1996 104 741 803 8880001
Kelce WR Stone CR Laws SC Gray LE Kemppainen JA Wilson EM Persistent DDT metabolite p,p'-DDE is a potent androgen receptor antagonist Nature 1995 375 581 585 7791873 10.1038/375581a0
Danzo BJ Environmental xenobiotics may disrupt normal endocrine function by interfering with the binding of physiological ligands to steroid receptors and binding proteins Environ Health Perspect 1997 105 294 301 9171990
Bonefeld-Jorgensen EC Andersen HR Rasmussen TH Vinggaard AM Effect of highly bioaccumulated polychlorinated biphenyl congeners on estrogen and androgen receptor activity Toxicology 2001 158 141 153 11275356 10.1016/S0300-483X(00)00368-1
Toft G Hagmar L Giwercman A Bonde JP The epidemiological evidence on reproductive effects of biopersistent organochlorines in humans Reprod Toxicol 2004 19 5 26 15336708 10.1016/j.reprotox.2004.05.006
Toft G Axmon A Giwercman A Thulstrup AM Rignell-Hydbom A Sloth Pedersen H Ludwicki JK Zvyesday V Spano M Manicardi GC Bonefeld-Jorgensen EC Hagmar L Bonde JP INUENDO Fertility in four regions spanning large contrasts in serum levels of widespread persistent organochlorines Environmental Health 2005 4 26 16280075 10.1186/1476-069X-4-26
Arctic Monitoring and Assessment Programme: Human Health in the Arctic, Arctic Monitoring and Assessment Programme (AMAP), Oslo 2003
Svensson B-G Nilsson A Jonsson E Schütz A Åkesson B Hagmar L Fish consumption and exposure to persistent organochlorine compounds, mercury, selenium and methylamines in Swedish fishermen Scand J Work Environ Health 1995 21 94 103
Rylander L Hagmar L Mortality and cancer incidence among women with a high consumption of fatty fish contaminated with persistent organochlorine compounds Scand J Work Environ Health 1995 21 419 426 8824747
Czaja K Ludwicki JK Góralczyk K Struciński P Organochlorine pesticides, HCB, and PCBs in human milk in Poland Bull Environ Contam Toxicol 1997 58 769 775 9115141 10.1007/s001289900400
Gladen BC Shkiryak-Nyzhnyk ZA Chyslovska N Zadorozhnaja TD Little RE Persistent organochlorine compounds and birth weight Ann Epidemiol 2003 13 151 157 12604157 10.1016/S1047-2797(02)00268-5
Grimvall W Rylander L Nilsson-Ehle P Nilsson U Strömberg U Hagmar L Östman C Monitoring of polychlorinated biphenyls in human blood plasma with respect to age, lactation and fish consumption; methodology developments Arch Environ Contam Toxicol 1997 32 329 336 9096084 10.1007/s002449900193
Glynn AW Wolk A Aune M Atuma S Zettermark S Maehle-Schmid M Darnerud PO Becker W Vessby B Adami HOl Serum concentrations of organochlorines in men: a search for markers of exposure Sci Total Environ 2000 263 197 208 11194153 10.1016/S0048-9697(00)00703-8
Deutch B Pedersen HS Jorgensen EC Hansen JC Smoking as a determinant of high organochlorine levels in Greenland Arch Environ Health 2003 58 30 36 12747516
Muckle G Ayotte P Dewailly E Jacobson SW Jacobson JL Determinants of polychlorinated biphenyls and methylmercury exposure in Inuit women of childbearing age Environ Health Perspect 2001 109 957 963 11673127
Rignell-Hydbom A Rylander L Giwercman A Jönsson BAG Nilsson-Ehle P Hagmar L Exposure to CB-153 and p,p'-DDE and male reproductive function Human Reprod 2004 19 2066 2075 10.1093/humrep/deh362
Rylander L Nilsson-Ehle P Hagmar L A simplified precise method for adjusting serum levels of persistent organohalogen pollutants to total serum lipids Chemosphere 2005 Jul 7
Altman DG Practical statistics for medical research 1991 London: Chapman and Hall
Bjerregaard P Dewailly E Ayotte P Pars T Ferron L Mulvad G Exposure of Inuit in Greenland to organochlorines through the marine diet J Toxicol Environ Health A 2001 62 69 81 11209822 10.1080/009841001455490
Falk C Hanrahan L Anderson HA Kanarek MS Draheim L Needham L Patterson D Jr Body burden levels of dioxin, furans, and PCBs among frequent consumers of Great Lakes sport fish. The Great Lakes Consortium Environ Res 1999 80 S19 S25 10092416 10.1006/enrs.1998.3906
Hunter DJ Hankinson SE Laden F Colditz GA Manson JE Willett WC Speizer FE Wolff MS Plasma organochlorine levels and the risk of breast cancer N Engl J Med 1997 337 1253 1258 9345073 10.1056/NEJM199710303371801
Rylander L Dyremark E Strömberg U Östman C Hagmar L The impact of age, lactation and dietary habits on PCB in plasma in Swedish women Sci Total Environ 1997 207 55 61 9397600 10.1016/S0048-9697(97)00245-3
Wolff MS Zeleniuch-Jacquotte A Dubin N Toniolo P Risk of breast cancer and organochlorine exposure Cancer Epidemiol Biomark Prev 2000 9 271 7
Laden F Neas LM Spiegelman D Hankinson SE Willett WC Ireland K Wolff MS Hunter DJ Predictors of plasma concentrations of DDE and PCBs in a group of U.S. women Environ Health Perspect 1999 107 75 81 9872720
Rogan WJ Gladen BC McKinney JD Carreras N Hardy P Thullen J Tingelstad J Tully M Polychlorinated biphenyls (PCBs) and dichlorodiphenyl dichloroethene (DDE) in human milk: effects of maternal factors and previous lactation Am J Public Health 1986 76 172 177 3080910
Skaare JU Polder A Polychlorinated biphenyls and organochlorine pesticides in milk of Norwegian women during lactation Arch Environ Contam Toxicol 1990 19 640 645 2122815 10.1007/BF01183978
Hong CS Xiao J Casey AC Bush B Fitzgerald EF Hwang SA Mono-ortho- and non-ortho-substituted polychlorinated biphenyls in human milk from Mohawk and control women: effects of maternal factors and previous lactation Arch Environ Contam Toxicol 1994 27 431 437 7944559 10.1007/BF00213183
Bignert A Olsson M Persson W Jensen S Zakrisson S Litzén K Eriksson U Haggberg L Alsberg T Temporal trends of organochlorines in Northern Europe, 1967–1995. Relation to global fractionation, leakage from sediments and international measures Environ Pollut 1998 99 177 198 15093312 10.1016/S0269-7491(97)00191-7
Glynn AW Granath F Aune M Atuma S Darnerud PO Bjerselius R Vainio H Weiderpass E Organochlorines in Swedish women: determinants of serum concentrations Environ Health Perspect 2003 111 349 355 12611665
Schildkraut JM Demark-Wahnefried W DeVoto E Hughes C Laseter JL Newman B Environmental contaminants and body fat distribution Cancer Epidemiol Biomark Prev 1999 8 179 83
Wolff MS Berkowitz GS Brower S Senie R Bleiweiss IJ Tartter P Pace B Roy N Wasllenstein S Weston A Organochlorine exposures and breast cancer risk in New York City women Environ Res 2000 84 151 161 11068929 10.1006/enrs.2000.4075
Dorgan JF Brock JW Rothman N Needham LL Miller R Stephenson HE JrSchussler N Taylor PR Serum organochlorine pesticides and PCBs and breast cancer risk: results from a prospective analysis (USA) Cancer Causes Control 1999 10 1 11 10334636 10.1023/A:1008824131727
Wolff MS Anderson HA Correspondence re: J. M. Schildkraut et al., Environmental contaminants and body fat distribution Cancer Epidemiol Biomark Prev 1999 8 951 952
Deutch B Hansen JC High human plasma levels of organochlorine compounds in Greenland. Regional differences and lifestyle effects Dan Med Bull 2000 47 132 137 10822803
Pereg D Dewailly E Poirier GG Ayotte P Environmental exposure to polychlorinated biphenyls and placental CYP1A1 activity in Inuit women from northern Quebec Environ Health Perspect 2002 110 607 612 12055053
Deutch B Pedersen HS Hansen JC Dietary composition in Greenland plasma fatty acids and persistent organic pollutants Sci Total Environ 2000 331 177 188 15325148
Dewailly E Ryan JJ Laliberté C Bruneau S Weber J-P Gingras S Carrier G Exposure of remote maritime populations to coplanar PCBs Environ Health Perspect 1994 102 205 209 8187710
Arctic Monitoring and Assessment Programme: Assessment report: Arctic pollution issues 1998 xii Oslo, Norway 859
Bjerregaard P Hansen JC Organochlorines and heavy metals in prgnant women from the Disko Bay area in Greenland Sci Tot Environ 2000 245 195 202 10.1016/S0048-9697(99)00444-1
Noren K Contemporary and retrospective investigations of human milk in the trend studies of organochlorine contaminants in Sweden Sci Total Environ 1993 139–140 347 355 10.1016/0048-9697(93)90032-2
Schade G Heinzow B Organochlorine pesticides and polychlorinated biphenyls in human milk of mothers living in northern Germany: current extent of contamination, time trend from 1986 to 1997 and factors that influence the levels of contamination Sci Total Environ 1998 215 31 39 9599454 10.1016/S0048-9697(98)00008-4
Waliszewski SM Aguirre AA Infanzon RM Rivera J Infanzon R Time trend of organochlorine pesticide residues in human adipose tissue in Veracruz, Mexico: 1988–1997 survey Sci Total Environ 1998 221 201 204 9842747 10.1016/S0048-9697(98)00279-4
Harris CA O'Hagan S Merson GH Organochlorine pesticide residues in human milk in the United Kingdom 1997-8 Hum Exp Toxicol 1999 18 602 606 10557010 10.1191/096032799678839392
Noren K Meironyte D Certain organochlorine and organobromine contaminants in Swedish human milk in perspective of past 20–30 years Chemosphere 2000 40 1111 23 10739053 10.1016/S0045-6535(99)00360-4
He JP Stein AD Humphrey HE Paneth N Courval JM Time trends in sport-caught Great Lakes fish consumption and serum polychlorinated biphenyl levels among Michigan Anglers, 1973–1993 Environ Sci Technol 2001 35 435 440 11351711 10.1021/es001067p
Dallaire F Dewailly É Laliberte C Muckle G Ayotte P Temporal trends of organochlorine concentrations in umbilical cord blood of newborns from the lower north shore of the St. Lawrence river (Québec, Canada) Environ Health Perspect 2002 110 835 838 12153768
Dallaire F Dewailly É Muckle G Ayotte P Time trends of persistent organic pollutants and heavy metals in umbilical cord blood of Inuit infants born in Nunavik (Québec, Canada) between 1994 and 2001 Environ Health Perspect 2003 111 1660 1664 14527847
Asplund L Svensson BG Nilsson A Eriksson U Jansson B Jensen S Wideqvist U Skerfving S Polychlorinated biphenyls, 1,1,1-trichloro-2,2-bis(p-chlorophenyl)ethane (p,p'-DDT) and 1,1-dichloro-2,2-bis(p-chlorophenyl)-ethylene (p,p'-DDE) in human plasma related to fish consumption Arch Environ Health 1994 49 477 486 7818291
Sjödin A Hagmar L Klasson-Wehler E Björk J Bergman Å Influence of the consumption of fatty Baltic Sea fish on plasma levels of halogenated environmental contaminants in Latvian and Swedish men Environ Health Perspect 2000 108 1035 1040 11102293
|
16283941
|
PMC1308838
|
CC BY
|
2021-01-04 16:36:32
|
no
|
Environ Health. 2005 Nov 11; 4:27
|
utf-8
|
Environ Health
| 2,005 |
10.1186/1476-069X-4-27
|
oa_comm
|
==== Front
Epidemiol Perspect InnovEpidemiologic perspectives & innovations : EP+I1742-5573BioMed Central London 1742-5573-2-101630955610.1186/1742-5573-2-10Analytic PerspectiveChoosing an appropriate bacterial typing technique for epidemiologic studies Foxman Betsy [email protected] Lixin [email protected] James S [email protected] Shannon D [email protected] Carl F [email protected] Department of Epidemiology and Center for Molecular and Clinical Epidemiology of Infectious Diseases, University of Michigan School of Public Health, Ann Arbor, Michigan, USA2 National Food Safety/ & Toxicology Center, Michigan State University, East Lansing, Michigan, USA2005 25 11 2005 2 10 10 13 6 2005 25 11 2005 Copyright © 2005 Foxman et al; licensee BioMed Central Ltd.2005Foxman et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
A wide variety of bacterial typing systems are currently in use that vary greatly with respect to the effort required, cost, reliability and ability to discriminate between bacterial strains. No one technique is optimal for all forms of investigation. We discuss the desired level of discrimination and need for a biologic basis for grouping strains of apparently different types when using bacterial typing techniques for different epidemiologic applications: 1) confirming epidemiologic linkage in outbreak investigations, 2) generating hypotheses about epidemiologic relationships between bacterial strains in the absence of epidemiologic information, and 3) describing the distributions of bacterial types and identifying determinants of those distributions. Inferences made from molecular epidemiologic studies of bacteria depend upon both the typing technique selected and the study design used; thus, choice of typing technique is pivotal for increasing our understanding of the pathogenesis and transmission, and eventual disease prevention.
molecular epidemiologymethodsbacteriatyping
==== Body
Introduction
Ever since Koch discovered how to grow bacteria in pure culture, the laboratory has been an integral component of epidemiologic studies of bacterial diseases. Over time, our ability to discriminate among bacterial strains from the same species has increased, enhancing outbreak investigations and surveillance, studies of the natural history of infection, and our understanding of the transmission, pathogenesis and phylogeny of bacteria.
Analysis
Bacterial typing systems
Traditional typing systems for discriminating between bacteria from a single species have been based on phenotype, such as serotype, biotype, phage typing, or antibiogram (susceptibility to one or more antibiotics). More recently, techniques have been developed based on indirect measures of genetic sequence (such as pulsed-field gel electrophoresis (PFGE)) and direct measures of genetic sequence (such as multilocus sequence typing (MLST)). Sequencing an entire bacterial genome, and, using microarray technologies, comparing strains to a reference strain (comparative genomic hybridization) is now technically feasible; however, the cost and time required limits the applicability for most epidemiologic studies. For example, in 2005, total genomic sequencing costs roughly 100 to 500 times more per strain than comparative hybridization (~$100,000 to $500,000 versus ~$1000 to $2000), and MLST (~$140) is quite costly compared to PFGE (~$20). Further, we have yet to characterize the range of variability among bacterial strains of a single species by various techniques, and thus lack an appropriate context for interpreting the observed variation.
Understanding the strengths and weaknesses of the chosen bacterial typing technique enhances interpretation and generalization of study results. A summary of common typing techniques and the relative discriminatory power, repeatability (same test result, given random error, for same analysis on same sample in the same laboratory), reproducibility (same test result, given random error, for same analysis on same sample in a different laboratory), timing and cost is presented in Table 1; techniques have been recently reviewed elsewhere [1-3]. We have ordered techniques from those with the highest to lowest discriminatory power, that is, ability to distribute strains into the greatest number of groups. Thus, if the entire genome of a bacteria is sequenced we will be able to detect even very small differences between strains, for example, changes in gene sequence that do not cause changes in the expressed proteins, such as point mutations that naturally occur over time as the bacteria divides. Common typing techniques used in epidemiologic studies sequence one or more genetic regions, for example multi-locus sequence typing (MLST), or use enzymes to cut part or all of the genome into pieces, for example, pulsed-field gel electrophoresis. The number and size of the pieces correspond to the number and location of restriction sites cut by the enzymes, and thus are an indirect measure of sequence. Other common techniques use the polymerase chain reaction targeted to specific sequences, for example ERIC-PCR; the resulting reactions yield fragments of different sizes, which can be used to discriminate between bacterial types. Generally speaking, sequence-based methods are most repeatable and reproducible. Gel-based methods are less so, because of the inherent variability of the technique [2,3].
Table 1 Comparison of Common Bacterial Typing Techniques by Relative Discriminatory Power, Reproducibility, Repeatability, and Whether They Give Information on Dispersed or Focal Parts of the Genome, Time Required and Cost
Typing Technique Relative discriminatory power Relative repeatability Relative reproducibility Dispersed or focal parts of the genome* Days required post culture Relative Cost** Notes
Sequencing of entire genome High High High Entire genome Months to years Very high
Comparative hybridization against array containing entire gene sequence High Medium to high Medium to high Dispersed Weeks to months High Microarrays are increasingly available for human pathogens – not all genes will be present in the sequenced strain
Direct sequencing of one or more genetic regions Moderate to high (depends on gene choice) High High Focal if only one region 2–3 Equipment: Medium to High
Labor & Supplies: Medium to High Initial selection of target genes might be time consuming.
Multilocus sequence typing (MLST) Moderate to high (depends on gene choice) High High Dispersed 3+ Equipment: Medium to High
Labor & Supplies: High Initial selection of target genes might be time consuming. Species specific.
Binary typing (presence/absence of selected genes or alleles across the genome) Moderate to high (depends on gene choice) High Potentially High Dispersed (if chose different genes across the genome) 2–3 Equipment: medium
Labor & Supplies: Medium Reliability dependent on DNA yield and purity
Pulsed-field gel electrophoresis (PFGE) Moderate to high (depends on number of bands observed) Medium=> High (depending on species) Medium =>High Dispersed 3 Equipment: High
Labor & Supplies: High Discrimination depends on type and number of enzymes selected.
Restriction fragment length polymorphism (RFLP) Moderate to High (depends on number of bands observed) Medium=>High Medium Dispersed 1–3 Medium
Amplification of a single target gene specific to a pathogen Moderate to high (depends on gene choice) High Medium=>High Focal <1 Equipment: Low to Medium
Labor & Supplies: Low
Amplified fragment length polymorphism (AFLP) Moderate to high High Medium=>High Dispersed 2 Equipment: Low to Medium
Labor & Supplies: Low
Automated ribotyping Moderate High High Focal 1 Equipment: High
Labor & Supplies: High Works for most bacterial species
Ribosomal RNA gel electrophoresis Moderate High High Focal 1 Equipment: Low
Labor & Supplies: Medium
Targeting known repetitive gene sequences (enterobacterial repetitive intergenic consensus sequences (ERIC), repetitive extragenic palindromic sequences (REP), DRE (double repetitive element), BOX, insertional sequence (IS), polymorphic GC-rich repetitive sequences (PGRS)) Low to moderate Medium Low Generally dispersed 1 Equipment: Low to Medium
Labor & Supplies: Low Patterns vary with equipment used
Random primers (randomly amplified polymorphic DNA (RAPD), arbitrary primed PCR (AP-PCR)) Low to moderate Low Low Dispersed 1 Equipment: Low to Medium
Labor & Supplies: Low Patterns vary with equipment used
Restriction endonuclease on a single amplified product Low to moderate (depends on amplicon) High High Focal 1–2 Equipment: Low to Medium
Labor & Supplies: Low
Plasmid profiles Low High Medium Focal 1 Equipment: Low
Labor & Supplies: Low
*Focal corresponds to interrogating a single loci. Dispersed means multiple loci are interrogated.
**Per isolate costs in US dollars in 2005, assuming all equipment are available, and the investigator has access to automatic sequencing, for PCR reactions are ~$5, PFGE~$20, MLST ~$140, comparative hybridization~$1000 to $2000 and total genomic sequencing (assuming a strain has already been sequenced)~$100,000 to $500,000.
Note: For a summary and details of these techniques, and assessments of repeatability and reproducibility, see Tenover, 1997 [1], Gurtler and Mayall 2001 [2] and VanBelkum, 2003 [3]. In general, sequence-based methods are most repeatable and reproducible. Gel-based methods are less so, because of the inherent variability of the technique.
Our intention is not to focus on a particular technique, as the techniques continue to change rapidly. Instead, we discuss the strengths and weaknesses of current bacterial typing techniques for particular epidemiologic applications, and provide some insight into what characteristics a typing technique should have when applied to a specific research question. We recognize that choice of a molecular tool is often up to laboratory personnel and not the epidemiologist; however, laboratorians are not always involved in study design or the interpretation of study results (although this is highly desirable). A laboratorian, whose expertise is in a particular typing technique, cannot be expected to give appropriate advice if s/he does not understand the research question asked. Similarly, an epidemiologist cannot appropriately analyze and interpret results of a typing technique if s/he does not understand what it is measuring. Furthermore, if there is a mismatch between typing technique and research question, the study results are less likely to answer the research question. Unfortunately, epidemiologists and laboratorians often have little training in each other's fields, do not share a common vocabulary, and have very different research perspectives. Thus, our goal is to provide guidance for the epidemiologist about working collaboratively with laboratories to choose the appropriate bacterial typing technique, and for interpreting the results.
Epidemiologic Applications of Bacterial Typing Techniques
Discriminatory power is the average probability that a typing system will assign the same strain type to strains randomly sampled from the same group. In a typical analysis, epidemiologists use questionnaire data to discriminate between groups. For example, if investigating a foodborne outbreak associated with a picnic, then the variable 'ate food at the picnic' will be a poor discriminator of disease risk (as probably all ate), but 'ate potato salad' or even 'ate potatoes' might accurately classify individuals into high and low risk groups (if an ingredient in the potato salad, such as the eggs or mayonnaise, was the culprit). If we classify individuals into groups by all variables measured simultaneously (e.g., age, gender, food preferences, medical history, etc.), then our measure will be highly discriminatory (as each individual might fall into a separate group) – although not necessarily informative with respect to disease risk. Thus, the most discriminatory grouping is not necessarily the most informative, particularly if the groupings are not associated with the outcome of interest.
Bacterial typing techniques are analogous, but may or may not provide an appropriately discriminatory grouping (similar to 'ate potato salad'). We have identified three purposes where molecular typing techniques are applied in epidemiologic studies (Table 2). We give an example of a research goal that relates to each purpose, provide an assessment of the required discriminatory power and need to infer genetic relationships and/or population structure for that particular application. Each purpose is discussed, in turn, below.
Table 2 Required Discriminatory Power and Need to Infer Genetic Relationships and/or Population Structure for Various Epidemiologic Applications of Bacterial Typing Techniques
Purpose Example Research Goal Discriminatory Power Needed Need to infer genetic relationships and/or population structure
Confirm epidemiologic linkage a. Determine if epidemiologically related cases share the identical organism. Result: either support or refute epidemiologic data. Low Low
Generate hypotheses about epidemiologic relationships between bacterial strains in the absence of epidemiologic data a. Determine if time-space clustering surveillance isolates have identical or related genetic types. Result: trigger further epidemiologic investigation of related isolates.
b. Determine if outbreak is propagated. Result: trigger investigation into how is spread and/or control actions to stop spread.
c. Relate clinical outcomes to strain types or to the presence of transferable genetic material, e.g., antimicrobial resistance on a plasmid. Result: improve patient care. Moderate to High Moderate
Describe distribution of bacterial types and identify the determinants of that distribution a. Test the hypothesis of clonal spread versus independent origin of a particular strain over disparate geographic areas. Result: Better predict emergence and spread of disease.
b. Determine flow of infection from one group to another. Result: Public health intervention
c. Identification of pathogenic factors. Result: Develop new interventions or therapies specific to those factors Moderate to High High
First, however, we wish to point out that bacterial typing is not always the correct classification tool, as outbreaks are not always caused by a single, virulent clone. Contamination of the water or food supply by sewage can lead to an outbreak of diarrhea caused by a variety of different agents [4-6] although clonal outbreaks also occur following sewage contamination [7]. Other examples are the breakdown of abattoir procedures that lead to contamination from cows colonized with diverse agents, or of nursery hygiene procedures allowing transmission from visitors to children.
Further, strain typing results must be interpreted in the context of epidemiologic evidence as well as the characteristics of the bacteria. Neither laboratory nor epidemiologic evidence is definitive, but each validates the other. When epidemiologic evidence suggests contamination arising from diverse sources, stricter molecular typing criteria should not be used to classify cases as epidemic related. If typing data suggests a high degree of similarity, epidemiologic evidence should be sought relevant to a single contamination episode.
Confirm Epidemiologic Linkage
One of the most common applications of bacterial typing in an epidemiologic study is in the context of an outbreak investigation. Bacterial typing is used to confirm or refute epidemiologic evidence that cases are linked or that a particular food item, water source, or fomite was the source of infection. In this situation the laboratory data is essentially confirmatory and the required discriminatory power and need to infer genetic relationships or structure is low. If there is strong epidemiologic evidence linking a specific food item with disease (common or point source), for example, we often make public health decisions based on that evidence alone – even if there is no supporting laboratory evidence. In the vast majority of foodborne outbreaks, the suspected food is not available for culture and a definitive linkage cannot be demonstrated [8]. Nonetheless, these investigations often successfully identify correctable breaks in hygiene practice. However, even modestly discriminatory techniques are useful since the laboratory evidence confirms the epidemiologic findings. For this type of confirmation, using a rapid and inexpensive technique (like ERIC-PCR) might be preferred since the cost and time associated with a more definitive technique (like MLST) would add little to our understanding of the source of infection or the ultimate policy decision.
Generate hypotheses about epidemiologic relationships between bacterial strains in the absence of epidemiologic data
Molecular typing has increased the power of surveillance data to detect outbreaks. The Foodborne Diseases Active Surveillance Network (FoodNet) conducted by the Centers for Disease Control and Prevention uses pulsed-field gel electrophoresis to type surveillance isolates for several foodborne pathogens, including E. coli O157:H7, nontyphoidal Salmonella serotypes, Listeria monocytogenes and Shigella [9]. Bacterial typing of space-time clusters has identified unsuspected linkages triggering investigations, as well as demonstrating that apparent clusters were not related, ruling out need for investigation [10].
Molecular typing also facilitates the detection of chains of transmission. Molecular typing led to a reassessment of the epidemiology of tuberculosis in the United States by establishing that tuberculosis does not require prolonged contact but can be transmitted in casual settings [11]. Typing also allows us to relate clinical outcome to strain types, distinguishing recent tuberculosis infection from reactivation of disease, [12] and establishing that an individual can be infected with a second, different tuberculosis strain following initial infection [13].
When the investigator needs to identify potential outbreaks by typing surveillance isolates, or to distinguish between point source and propagated outbreaks, a more discriminatory technique is required. In a common or point source outbreak we expect the causative agent to be similar in all infected persons. Therefore, a more discriminatory technique is necessary to determine if a space-time cluster of isolates detected via surveillance represents a potential outbreak compared to a technique for typing isolates already epidemiologically linked. In a propagated outbreak or when tracking chains of transmission, the genetic sequence of the bacteria may be slightly different at the end compared to the beginning of the outbreak (how fast this occurs depends on the bacteria, however). If the bacteria are naturally competent, i.e., easily uptake DNA from other members of the species, such as non-typeable Haemophilus influenzae [14], a highly discriminatory typing technique may erroneously misclassify epidemic cases identified at the end of the epidemic as non-epidemic, particularly if there are no endemic strains available for comparison. Using a typing technique that allows classification consistent with phylogenetic relationships (e.g., MLST), or, if the bacteria is highly recombinant, with clonal complexes, is helpful as there is a biologically meaningful way to group strains (that is, logically collapse groups of related strains). Unfortunately, many typing techniques are analogous to nominal scales, e.g., ERIC: the groups are different from each other, but we cannot say which of the identified groups are more similar than others. Even for PFGE, which can be used to assess relatedness, similarity may vary by choice or number of restriction enzymes used. Further, the published criteria for PFGE relatedness (based on number of matching bands) were intended solely for outbreak situations and when isolates were collected over a short time period (<1 year) and there is an implied epidemiologic linkage [15].
Describe distribution of bacterial types and identify the determinants of that distribution
Advances in molecular genetics have facilitated the description of the genetic diversity of bacterial populations. Molecular genetic techniques have been used to distinguish if there have been independent spontaneous mutations leading to antibiotic resistance or if resistance was transmitted between strains via a mobile genetic element. In other applications molecular genetic techniques have determined the flow of infection from one group to another. These descriptive molecular epidemiologic studies often use strains collected from disparate areas and the epidemiologic and clinical information is minimal or non-contributory. In this case the chosen bacterial typing technique must be interpretable in terms of genetic distance (phylogeny) for the given time period and organism. Further, the technique should reflect whether the hypothesis is of clonal spread of a strain or of a mobile genetic element, (e.g., plasmid).
Some typing techniques are based on conserved genes within the bacterial genome, e.g., genes associated with metabolism or other 'housekeeping' functions, and others on more variable genes, e.g., genes associated with virulence. On average, when bacterial strains are compared using a genetic typing technique, there are fewer genetic differences between bacterial strains in the conserved genes than variable genes. Thus, typing techniques based on differences in conserved genes, such as MLST, will place strains into fewer, larger, groups, than typing techniques based on more variable genes, such as PFGE. Put another way, PFGE is generally more discriminatory than MLST.
For bacterial characteristics that are dependent both on the conserved and variable portions of the genome, such as virulence, the use of multiple typing techniques may be helpful, see, for example, [16]. Selection of the appropriate typing technique and a valid interpretation of the results for studies of distribution of bacterial types and the determinants of that distribution is easiest when at least some preliminary data are available. For example, knowledge of the rarity of the observed groups in the community, propensity of the species to acquire insertion elements or phage, the timing of strain collection and the evolutionary clock of the organism, that is, how quickly mutations occur or horizontal elements are acquired provides important information for both technique selection and interpretation of resulting findings.
The identification of pathogenic factors is an exercise in identifying what is different between strains causing and not causing disease. This identification proceeds in the manner of a case-control study with the bacterial agent as the unit of analysis [see, for example, [17]]. Standard epidemiologic study design issues apply: the study population must include both disease-causing and commensal isolates. Most disease-causing strains will predominate in a culture; non-pathogenic, or commensal organisms are often comprised of a mixture of strains of the same species. The investigator must select isolates for study accordingly. For example, E. coli is a common bowel inhabitant and is also the most common cause of urinary tract infection. Typically an individual has several E. coli strains in the bowel flora but urinary tract infection among outpatients is almost always caused by a single strain. The investigator must decide if the predominant isolate in the bowel flora is the one of interest or if several isolates should be selected for testing. If the objective were to link the bowel to the urinary tract flora, then choosing only the predominant bowel strain would not be sufficient. Identifying common elements generating pathogenicity may be the study objective: when the typing technique is unable to discriminate between pathogenic and diverse commensal isolates, epidemiologic and clinical information should be used to make that distinction, such as grouping together E. coli that cause urinary tract infection.
Pathogenicity determinants are often present on transferable genetic material, such as plasmids, pathogenicity islands, phages, etc. Transferable genetic material has a genetic history distinct from the rest of the host bacterial genome. In this case, phylogenetic analyses of these elements can provide useful information. For example, pathogenicity islands (PAIs) have been associated with a variety of conditions, including diarrhea and urinary tract infection [18-20]; specific virulence factor genes found on the PAIs encode for proteins that contribute directly to disease.
Conclusion
The application and interpretation of bacterial typing tools in epidemiologic studies requires understanding of both the strengths and limitations of the chosen bacterial typing technique as well as the epidemiologic study design to answer the research question. Beyond standard reliability, validity and cost considerations, key characteristics of a typing technique are 1) the ability to discriminate between strains and 2) a biologic basis for grouping strains with apparently different types. The level of discrimination required and need to be able to group strains depends on the research question. Similar to the desirability of including a statistician in the design phase so that the study design will result in appropriate data for the desired analysis, integrating an expert in the different typing techniques during the design phase will improve how well the research protocol fits the question(s) of interest.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
BF took the lead on drafting the manuscript, LZ took the lead on Table 1, and JSK outlined Table 2. All authors contributed to discussions leading to the manuscript, critiqued multiple drafts and approved the final manuscript.
Acknowledgements
The authors thank the members of the Center for Molecular and Clinical Epidemiology of Infectious Diseases faculty discussion group for their insights into this topic. This work was supported by RO1 DK35368 (BF), R21 AI44868 (BF) and R01 DK 55496 (CFM).
==== Refs
Tenover FC Arbeit RD Goering RV the Molecular Typing Working Group of the Society for Healthcare Epidemiology of America How to select and interpret molecular typing methods for epidemiological studies of bacterial infections: a review for healthcare epidemiologists Infect Control Hosp Epidemiol 1997 18 426 39 9181401
Gurtler V Mayall BC Genomic approaches to typing, taxonomy and evolution of bacterial isolates Int J Syst Evol Microbiol 2001 51 3 16 11211268
van Belkum A High-throughput epidemiologic typing in clinical microbiology Clin Microbiol Infect 2003 9 86 100 12588328 10.1046/j.1469-0691.2003.00549.x
Kapadia CR Bhat P Baker SJ Mathan VI A common-source epidemic of mixed bacterial diarrhea with secondary transmission Am J Epidemiol 1984 120 743 9 6093502
Berkelman RL Cohen ML Yashuk J Barrett T Wells JG Blake PA Traveler's diarrhea at sea: two multi-pathogen outbreaks caused by food eaten on shore visits Am J Pub Health 1983 73 770 2 6859361
Starko KM Lippy EC Dominguez LB Haley CE Fisher HJ Campers' diarrhea outbreak traced to water-sewage link Pub Health Rep 1986 101 527 31 3094085
Rosenberg ML Koplan JP Wachsmuth IK Wells JG Gangarosa EJ Guerrant RL Sack DA Epidemic diarrhea at Crater Lake from enterotoxigenic Escherichia coli. A large waterborne outbreak Ann Intern Med 1977 86 714 8 326116
Bean NH Griffin PM Goulding JS Ivey CB Foodborne Disease Outbreaks, 5-Year Summary, 1983–1987 MMWR 39 15 23 March 01, 1990 2156148
Centers for Disease Control and Prevention PulseNet Home Page Accessed 10/13/05.
Bender JB Hedberg CW Besser JM Boxrud DJ MacDonald KL Osterholm MT Surveillance by molecular subtype for Escherichia coli O157:H7 infections in Minnesota by molecular subtyping N Engl J Med 1997 337 388 94 9241128 10.1056/NEJM199708073370604
Golub JE Cronin WA Obasanjo OO Coggin W Moore K Pope DS Thompson D Sterling TR Harrington S Bishai WR Chaisson RE Transmission of Mycobacterium tuberculosis through casual contact with an infectious case Arch Intern Med 2001 161 2254 8 11575983 10.1001/archinte.161.18.2254
Small PM Hopewell PC Singh SP Paz A Parsonnet J Ruston DC Schecter GF Daley CL Schoolnik GK The epidemiology of tuberculosis in San Francisco. A population-based study using conventional and molecular methods N Engl J Med 1994 330 1703 9 7910661 10.1056/NEJM199406163302402
van Rie A Warren R Richardson M Victor TC Gie RP Enarson DA Beyers N van Helden PD Exogenous reinfection as a cause of recurrent tuberculosis after curative treatment N Engl J Med 1999 341 1174 9 10519895 10.1056/NEJM199910143411602
Gilsdorf JR Marrs CF Foxman B H. influenzae virulence factors: epidemiology and diversity Infect Immun 2004 72 2457 61 15102751 10.1128/IAI.72.5.2457-2461.2004
Tenover FC Arbeit RD Goering RV Mickelsen PA Murray BE Persing DH Swaminathan B Interpreting chromosomal DNA restriction patterns produced by pulsed-field gel electrophoresis: criteria for bacterial strain typing J Clin Microbiol 1995 33 2233 39 7494007
Beres SB Sylva GL Sturdevant DE Granville CN Liu M Ricklefs SM Whitney AR Parkins LD Hoe NP Adams GJ Low DE DeLeo FR McGeer A Musser JM Genome-wide molecular dissection of serotype M3 group A Streptococcus strains causing two epidemics of invasive infections Proc Natl Acad Sci USA 101 11833 8 2004 Aug 10 15282372 10.1073/pnas.0404163101
Zhang L Foxman B Manning SD Tallman P Marrs CF Molecular epidemiologic approaches to UTI gene discovery in uropathogenic Escherichia coli Infect Immun 2000 68 2009 15 10722596 10.1128/IAI.68.4.2009-2015.2000
McDaniel TK Jarvis KG Donnenberg MS Kaper JB A genetic locus of enterocyte effacement conserved among diverse enterobacterial pathogens Proc Natl Acad Sci USA 1995 92 1664 68 7878036
Lee CA Pathogenicity islands and the evolution of bacterial pathogens Infect Agen Dis 1996 5 1 7
Hacker J Blum-Oehler G Muhldorfer I Tschape H Pathogenicity islands of virulent bacteria: structure, function and impact on microbial evolution Mol Microbiol 1997 23 1089 97 9106201 10.1046/j.1365-2958.1997.3101672.x
|
16309556
|
PMC1308839
|
CC BY
|
2021-01-04 16:36:38
|
no
|
Epidemiol Perspect Innov. 2005 Nov 25; 2:10
|
utf-8
|
Epidemiol Perspect Innov
| 2,005 |
10.1186/1742-5573-2-10
|
oa_comm
|
==== Front
Harm Reduct JHarm Reduction Journal1477-7517BioMed Central London 1477-7517-2-241628197910.1186/1477-7517-2-24Brief ReportPrevalence and correlates of abscesses among a cohort of injection drug users Lloyd-Smith Elisa [email protected] Thomas [email protected] Robert S [email protected] Kathy [email protected] Julio SG [email protected] Evan [email protected] Department of Health Care and Epidemiology, Faculty of Medicine; University of British Columbia, 5804 Fairview Ave, Vancouver, Canada2 British Columbia Centre for Excellence in HIV/AIDS; St. Paul's Hospital, 608-1081 Burrard Street, Vancouver, Canada3 Department of Medicine; University of British Columbia, 3300 – 950 W. 10th Ave, Vancouver, Canada2005 10 11 2005 2 24 24 20 7 2005 10 11 2005 Copyright © 2005 Lloyd-Smith et al; licensee BioMed Central Ltd.2005Lloyd-Smith et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Recent studies have indicated that injection-related infections such as abscesses and cellulitis account for the majority of emergency room visits and acute hospitalizations accrued by local injection drug users. The objective of this analysis was to examine the prevalence and correlates of developing an abscess among a cohort of injection drug users in Vancouver and to identify socio-demographic and drug use variables associated with abscesses at baseline. We examined abscesses among participants enrolled in a prospective cohort of injection drug users. Categorical variables were analyzed using the Pearson's chi-square test and continuous variables were analyzed using the Wilcoxon signed rank test. Among 1 585 baseline participants, 341 (21.5%) reported having an abscess in the last six months. In a logistic regression model that adjusted for all variables that were associated with having an abscess at p < 0.1 in univariate analyses, female gender [odds ratio (OR) = 1.7, [95%CI: 1.2 – 2.4]; p = 0.002), recent incarceration (OR = 1.7, [95%CI: 1.3 – 2.2]; p < 0.001), sex trade involvement (OR = 1.4 [95% CI: 1.0 – 2.0]; p = 0.03), frequent cocaine use (OR = 1.5 [95%CI: 1.2 – 2.0]; p = 0.002) and HIV serostatus (OR = 1.5, [95%CI: 1.2 – 2.0]; p = 0.003) were positively associated with having an abscess. Explanations for these associations require further study, and interventions are needed to address this highly prevalent concern.
==== Body
Findings
The Downtown Eastside of Vancouver, Canada is a community characterized by high rates of HIV among injection drug users (IDU), and is also the setting of one of North America's highest volume needle exchange program (NEP) [1]. Recent studies have indicated that injection-related infections, such as abscesses and cellulitis, account for the majority of emergency room visits and acute hospitalizations in local IDU [2,3]. Factors associated with the development of abscesses among IDU have not been well described in settings with widespread access to sterile injecting equipment and high rates of HIV infection. In particular, abscesses are not characterized in Vancouver. However, abscesses can lead to serious complications including but not limited to osteomyelitis [4], endocarditis [5-7], and septicemia [8,9]. An ongoing prospective cohort study of IDU in Vancouver allowed for an examination of the prevalence and factors associated having an abscess in this setting.
For these analyses, data was collected through the Vancouver Injection Drug Users Study (VIDUS), a prospective cohort that has been previously described in detail [1]. Data from participants who completed baseline questionnaires between May 1, 1996 and May 31, 2004 were evaluated for the present study. Participants were categorized on the basis of whether or not they reported having an abscess lasting for more than three days during the previous six months. Univariate and multivariable statistics were applied to determine factors associated with developing an abscess in the previous six months. Categorical variables were analyzed using the Pearson's chi-square test, and continuous variables were analyzed using the Wilcoxon signed rank test. Variables associated with having an abscess at p < 0.1 were considered in a subsequent logistic regression analysis.
Socio-demographic and drug-using characteristics considered in these analyses as potential risk factors included: age, gender, HIV status, unstable housing, residing in Vancouver's Downtown Eastside, incarceration in the previous six months, sex trade involvement, borrowing and lending of syringes, frequent heroin and cocaine injection, binge drug use, public drug injection, requiring help with injections, and methadone maintenance therapy use. Unstable housing was defined as living in a single room occupancy hotel, transitional living arrangements, or being homeless. Individuals who reported injecting cocaine or heroin once or more a day were defined as frequent heroin and cocaine injectors. Bingeing was defined as periods in which drugs were injected more often than usual. Variable definitions were consistent with previous analyses [1].
Overall, of the 1 585 baseline VIDUS participants, 341 (21.5%) reported having an abscess in the last six months. The factors associated with having an abscess at p < 0.1 in univariate analyses included: female gender (OR = 2.4, [95%CI: 1.8 – 3.0]; p < 0.001); unstable housing (OR = 1.3, [95%CI: 1.1 – 1.8]; p = 0.01); recent incarceration (OR = 1.7, [95%CI: 1.3 – 2.1]; p < 0.001); sex trade involvement (OR = 2.4, [95%CI: 1.9 – 3.1]; p < 0.001); frequent heroin use (OR = 1.4, [95%CI: 1.1 – 1.8]; p = 0.006); frequent cocaine use (OR = 1.9, [95%CI: 1.5 – 2.5]; p < 0.001); residing in Vancouver's Downtown Eastside (OR = 1.5, [95%CI: 1.1 – 1.9]; p = 0.003); and HIV serostatus (OR = 1.8, [95%CI: 1.4 – 2.3]; p < 0.001). Table 1 shows the baseline demographic characteristics of IDU stratified by having an abscess or not in the past six months for significant variables considered in the univariate analysis.
Table 1 Baseline demographic characteristics of IDU stratified by having an abscess in the past six months.
Characteristic No abscess past six months n = 1 244 Abscess past six months n = 341 Odds Ratio (95% CI) p-value
Gender
Male 848 (68.2) 162 (47.5)
Female 396 (31.8) 179 (52.5) 2.4 (1.9 – 3.0) < 0.001
HIV status
Negative 919 (73.9) 209 (61.3)
Positive 325 (26.1) 132 (38.7) 1.7 (1.4 – 2.3) < 0.001
Unstable housing*
No 492 (39.5) 109 (32.0)
Yes 752 (60.5) 387 (68.0) 1.3 (1.1 – 1.8) 0.011
Recent incarceration*
No 861 (69.2) 196 (57.5)
Yes 383 (30.8) 145 (42.5) 1.7 (1.3 – 2.1) <0.001
DTES residence*
No 552 (44.4) 121 (35.5)
Yes 692 (55.6) 220 (64.5) 1.5 (1.1 – 1.9) 0.003
Sex trade involved*
No 942 (75.7) 191 (56.0)
Yes 302 (24.3) 150 (44.0) 2.4 (1.9 – 3.1) < 0.001
Heroin use*
Less than daily 840 (67.5) 203 (59.5)
Daily use 404 (32.5) 138 (40.5) 1.4 (1.1 – 1.8) 0.006
Cocaine use*
Less than daily 858 (69.0) 182 (53.4)
Daily use 386 (31.0) 159 (46.6) 1.9 (1.5 – 2.5) < 0.001
Note: IDU = injection drug user, DTES = Downtown Eastside Residence. *Indicates behaviour during the six month period prior to the baseline interview.
As shown in Table 2, in a logistic regression model that adjusted for all variables that were associated with having an abscess at p < 0.1 in univariate analyses, female gender [odds ratio (OR) = 1.7, [95% CI: 1.2 – 2.4]; p = 0.002), recent incarceration (OR = 1.7, [95% CI: 1.3 – 2.2]; p < 0.001), sex trade involvement (OR = 1.4 [95% CI: 1.0 – 2.0]; p = 0.030), frequent cocaine use (OR = 1.5 [95% CI: 1.2 – 2.0]; p = 0.002) and HIV serostatus (OR = 1.5, [95% CI: 1.2 – 2.0]; p = 0.003) were independently and positively associated with having an abscess.
Table 2 Logistic regression of factors associated with having an abscess
Characteristic Odds Ratio 95% C.I. p-value
Gender
(Female vs Male) 1.7 (1.4 – 2.4) 0.002
Frequent cocaine use
(Yes vs No) 1.5 (1.2 – 2.0) 0.002
Recent incarceration
(Yes vs No) 1.7 (1.3 – 2.2) <0.001
Sex trade involvement
(Yes vs No) 1.5 (1.1 – 2.1) 0.030
HIV serostatus
(Yes vs No) 1.5 (1.2 – 2.0) 0.003
Our results indicate female gender, recent incarceration, sex trade involvement, frequent cocaine use and HIV serostatus are positively associated with developing an abscess. These results are consistent with results from a study in Amsterdam where female gender and prostitution among women, as well as, frequent cocaine use were identified as independently and positively associated with skin abscesses [10]. In addition, the association between HIV-positive status and having an abscess has been noted elsewhere, and is understandable given that HIV-positive individuals may have a heightened susceptibility to bacterial infections [11,12]. Furthermore, high risk of infectious complications, such as endocarditis from abscesses [10], occur among HIV infected individuals [11].
Abscesses are a common consequence of injection drug use [13-15]. The present study demonstrates that widespread access to sterile syringes through high-volume needle exchanges and a medically supervised safer injection facility may not be sufficient to prevent high rates of abscesses among IDU in Vancouver. In addition, our findings demonstrate the need for educational and structural interventions to improve rates of sterile injecting [16,17].
Our study has limitations. First, although previous research has indicated that the VIDUS cohort is representative of Vancouver IDU [18], VIDUS is not a random sample. Second, the study relied on self-report, and therefore, the results could be susceptible to socially desirable reporting although we know of no reason why reporting abscesses would be subject to this concern. Third, the cross-sectional nature of this study precludes any conclusions regarding causal relationships between the variables studied and the outcome of interest. Further prospective study is needed to assess predictors of abscess in this setting.
In summary, 21.5% of IDU participating in this study reported having an abscess in the previous six months. Results from this study indicate female gender, recent incarceration, sex trade involvement, frequent cocaine use and HIV serostatus are independently and positively associated with developing an abscess in injection drug users. Given the potential health complications arising from bacterial infections our findings highlight the need for the expansion of programs to promote safer injection practices.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Elisa Lloyd-Smith, Thomas Kerr and Evan Wood designed the study. Kathy Li conducted the statistical analysis. Elisa Lloyd-Smith drafted the manuscript and incorporated all suggestions. All coauthors made significant contributions to the conception and design of the analyses, interpretation of the data and drafting of the manuscript, and they all approved the version to be published.
Acknowledgements
We would particularly like to thank the VIDUS participants for their willingness to participate in the study. We also thank Drs. Kevin Craib, Richard Harrigan, Cari Miller, David Patrick, Mark Tyndall, Martin Schechter, Will Small, Patricia Spittal, & Steffanie Strathdee, for their research assistance, and Bonnie Devlin, John Charette, Caitlin Johnston, Vanessa Volkommer, Steve Kain, Dave Isham, Nancy Lalibarte, Sue Currie and Peter Vann for their administrative assistance. The study was supported by the US National Institutes of Health (R01 DA011591-04A1) and CIHR grant (MOP-67262). Elisa Lloyd-Smith is supported by a Junior Graduate Studentship from the Michael Smith Foundation for Health Research.
==== Refs
Wood E Tyndall MW Spittal PM Li K Hogg RS Montaner JS O'Shaughnessy MV Schechter MT Factors associated with persistent high-risk syringe sharing in the presence of an established needle exchange programme Aids 2002 16 941 943 11919503 10.1097/00002030-200204120-00021
Kerr T Wood E Grafstein E Ishida T Shannon K Lai C Montaner J Tyndall MW High rates of primary care and emergency department use among injection drug users in Vancouver J Public Health (Oxf) 2004 15564279
Palepu A Tyndall MW Leon H Muller J O'Shaughnessy MV Schechter MT Anis AH Hospital utilization and costs in a cohort of injection drug users Cmaj 2001 165 415 420 11531049
Roszler MH McCarroll KA Donovan KR Rashid T Kling GA The groin hit: complications of intravenous drug abuse Radiographics 1989 9 487 508 2727357
DeWitt DE Paauw DS Endocarditis in injection drug users Am Fam Physician 1996 53 2045 2049 8623717
DiNubile MJ Short-course antibiotic therapy for right-sided endocarditis caused by Staphylococcus aureus in injection drug users Ann Intern Med 1994 121 873 876 7978701
Miro JM Moreno A Mestres CA Infective Endocarditis in Intravenous Drug Abusers Curr Infect Dis Rep 2003 5 307 316 12866981
Hankins C Palmer D Singh R Unintended subcutaneous and intramuscular injection by drug users Cmaj 2000 163 1425 1426 11192645
Williamson N Archibald C Van Vliet JS Unexplained deaths among injection drug users: a case of probable Clostridium myonecrosis Cmaj 2001 165 609 611 11563214
Spijkerman IJ van Ameijden EJ Mientjes GH Coutinho RA van den Hoek A Human immunodeficiency virus infection and other risk factors for skin abscesses and endocarditis among injection drug users J Clin Epidemiol 1996 49 1149 1154 8826995 10.1016/0895-4356(96)00180-1
Selwyn PA Alcabes P Hartel D Buono D Schoenbaum EE Klein RS Davenny K Friedland GH Clinical manifestations and predictors of disease progression in drug users with human immunodeficiency virus infection N Engl J Med 1992 327 1697 1703 1359411
Flanigan TP Hogan JW Smith D Schoenbaum E Vlahov D Schuman P Mayer K Self-reported bacterial infections among women with or at risk for human immunodeficiency virus infection Clin Infect Dis 1999 29 608 612 10530455
Brown PD Ebright JR Skin and Soft Tissue Infections in Injection Drug Users Curr Infect Dis Rep 2002 4 415 419 12228028
Murphy EL DeVita D Liu H Vittinghoff E Leung P Ciccarone DH Edlin BR Risk factors for skin and soft-tissue abscesses among injection drug users: a case-control study Clin Infect Dis 2001 33 35 40 11389492 10.1086/320879
Binswanger IA Kral AH Bluthenthal RN Rybold DJ Edlin BR High prevalence of abscesses and cellulitis among community-recruited injection drug users in San Francisco Clin Infect Dis 2000 30 579 581 10722447 10.1086/313703
Ross MW Wodak A Stowe A Gold J Explanations for sharing injection equipment in injecting drug users and barriers to safer drug use Addiction 1994 89 473 479 8025506
Wood E Kerr T Montaner JS Strathdee SA Wodak A Hankins CA Schechter MT Tyndall MW Rationale for evaluating North America's first medically supervised safer injecting facility Lancet Infect Dis 2004 4 301 306 15120347 10.1016/S1473-3099(04)01006-0
Tyndall MW Craib KJ Currie S Li K O'Shaughnessy MV Schechter MT Impact of HIV infection on mortality in a cohort of injection drug users J Acquir Immune Defic Syndr 2001 28 351 357 11707672
|
16281979
|
PMC1308840
|
CC BY
|
2021-01-04 16:36:49
|
no
|
Harm Reduct J. 2005 Nov 10; 2:24
|
utf-8
|
Harm Reduct J
| 2,005 |
10.1186/1477-7517-2-24
|
oa_comm
|
==== Front
Head Face MedHead & Face Medicine1746-160XBioMed Central London 1746-160X-1-81627090810.1186/1746-160X-1-8ReviewPalatal development of preterm and low birthweight infants compared to term infants – What do we know? Part 1: The palate of the term newborn Hohoff Ariane [email protected] Heike [email protected] Ulrike [email protected] Erik [email protected] Poliklinik für Kieferorthopädie, Universitätsklinikum, Westfälische Wilhelms-Universität, Münster, Germany2 Department of Neonatology, Brighton & Sussex University Hospitals, UK3 Klinik für Kinderheilkunde, Division of Neonatology, Universitätsklinikum, Westfälische Wilhelms-Universität, Münster, Germany2005 28 10 2005 1 8 8 8 9 2005 28 10 2005 Copyright © 2005 Hohoff et al; licensee BioMed Central Ltd.2005Hohoff et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The evidence on prematurity as 'a priori' a risk for palatal disturbances that increase the need for orthodontic or orthognathic treatment is still weak. Further well-designed clinical studies are needed. The objective of this review is to provide a fundamental analysis of methodologies, confounding factors, and outcomes of studies on palatal development. One focus of this review is the analysis of studies on the palate of the term newborn, since knowing what is 'normal' is a precondition of being able to assess abnormalities.
Methods
A search profile based on Cochrane search strategies applied to 10 medical databases was used to identify existing studies. Articles, mainly those published before 1960, were identified from hand searches in textbooks, encyclopedias, reference lists and bibliographies. Sources in English, German, and French of more than a century were included. Data for term infants were recalculated if particular information about weight, length, or maturity was given. The extracted values, especially those from non-English paper sources, were provided unfiltered for comparison.
Results
The search strategy yielded 182 articles, of which 155 articles remained for final analysis. Morphology of the term newborn's palate was of great interest in the first half of the last century. Two general methodologies were used to assess palatal morphology: visual and metrical descriptions. Most of the studies on term infants suffer from lack of reliability tests. The groove system was recognized as the distinctive feature of the infant palate. The shape of the palate of the term infant may vary considerably, both visually and metrically. Gender, race, mode of delivery, and nasal deformities were identified as causes contributing to altered palatal morphology. Until today, anatomical features of the newborn's palate are subject to a non-uniform nomenclature.
Conclusion
Today's knowledge of a newborn's 'normal' palatal morphology is based on non-standardized and limited methodologies for measuring a three-dimensional shape. This shortcoming increases bias and is the reason for contradictory research results, especially if pathologic conditions like syndromes or prematurity are involved. Adequate measurement techniques are needed and the 'normal palatal morphology' should be defined prior to new clinical studies on palatal development.
==== Body
Background
Preterm infants, i.e. those born before completion of 37 gestational weeks, account for 6–10% of births in Western society [1-3]. Preterm infants form the majority of low birthweight infants [3]. By definition, neonates weighing less than 2500 g are described as low birthweight infants [4]. The proportion of neonates weighing less than 1500 g (very low birthweight) is approximally 1–1.5% of all newborns [3]. As preterm infants <1500 – 1800 g and <32nd – 34th gestational week have an insufficiently developed sucking response, they normally have to be fed through an orogastric tube.
The factors discussed as potential triggers of a premature birth include: high or low age of the mother, low socio-economic status, inadequate antenatal care, drug, alcohol and nicotine abuse, diabetes, multiple pregnancies [5], anemia, previous miscarriages or abortions, deformity of the uterus, abnormal presentation of the fetus, endocrine disorders, excessive mental or physical strain on the pregnant woman [6], stress [7], hypertension [8], and infections [9]. The potential influence of periodontal infections in the mother on the risk of a premature birth is a matter on which there is no general agreement: Many studies report an increased risk [10-14] whereas others found no evidence of such an association [15].
Preterm infants suffer not only from the effects of a shorter antenatal development period but from the immaturity of their organs [4], involving the risk of neonatal complications such as immaturity of the liver and kidney vitamin d metabolism [16,17], pulmonary diseases, hyperbilirubinemia and hypocalcemia [18] as well as respiratory distress, apnea, hypoglycemia, cardiac defects, infections, metabolic bone diseases and intracranial hemorrhages [19]. The latter have a significant impact on the function of rooting, non-nutritive sucking and suck-swallow responses [20].
Longitudinal studies confirm that the physical and cognitive development of most VLBW infants is delayed [21-24]. The median incidence of cerebral palsy is 7.7%, and that of disability 25% [23]. Real chances of survival without a substantially impaired state of health are not to be expected before completion of the 25th – 26th gestational week. Although substantial health impairment is to be expected in 2/3 to 3/4 of infants born in the 24th gestational week, the survival rate is meanwhile 50 – 60%, and in those born in the 25th GW as high as 70 – 80% [3].
With the survival prospects of preterm infants having undergone such a dramatic improvement [2,25-29], research into the development of these small patients can and must now be extended beyond securing their mere survival to other areas such as their physical and cognitive development. The morbidity potential associated with the premature birth needs to be investigated [30-32]. Only if the problems resulting from premature birth are exactly known preventive measures can be taken.
The orofacial region plays an important role in the infant's development in general: the mouth has been described as the 'cockpit of the awareness of the term infant and of its most discriminate responses' [33]. However, in the early stages of the development of the oral cavity, the soft bones of the palate are malleable and pressure from any object can easily mould the shape of the palate [34]. Thus, at an early stage the palate in particular may be subject to influences such as mode of delivery [35], positioning and gravitational forces [2], oral intubation, sucking respectively inadequate sucking response, delayed primary tooth eruption or general hypotonia and its development may in turn affect the infant's food intake, breathing, phonation, dental development, facial appearance [2], esthetics and psychosocial development (Figure 1a–e).
Figure 1 a-e Facial appearance (a, b), circular open bite (c) and palatal aspect (d) of a postnatally orotracheally intubated preterm infant in the initial phase of the dentition. Notice that the teeth of the child are 'in occlusion' on Figures a-c. The food intake – limited to soft or mashed foods due to the extreme dysgnathia – leads to marked frustration. The infant is teased because of its eating problems and the shape of its jaw and head. Radiography revealed premature ossification of the median suture (e).
The evidence of these consequences is still weak. A recent published systematic review [36] could not answer the questions on whether premature birth causes permanent alteration of palatal morphology, alteration of dental occlusion, and altered tooth-crown dimensions. The scientific evidence was too weak because of the contradictory results and lack of longitudinal studies.
Systematic reviews are not subject to the weakness of conventional narrative literature reviews because of the defined methods used to identify and reject studies; therefore, the conclusions are more reliable. However, systematic reviews are also open to questioning. The main issue to which criticism is addressed is the oversimplification of results by focusing on overall effects and downplaying mediating effects [37]. Contradictory results and methodological heterogeneity are common problems in the constitution of a systematic review. Six out of seven recently published systematic reviews (PubMed search: 'systematic review' AND orthod*) [36,38-43] concluded – irrespective of the research question – that further well-designed studies are needed.
It is therefore necessary to provide prospective investigators with methodological details of previous studies, especially in the field of morphometric assessment of palatal development, where new and more accurate methods have been established in the recent past. Moreover, information in different languages and without a restriction to particular databases and time periods must be included – a precondition which has not been considered yet.
The objective of this review is to provide a fundamental analysis of methodologies, confounding factors, and outcomes of studies on palatal development of preterm and low birthweight infants as compared to term infants. This review will be a major source of unfiltered data from more than a century, including also literature in German and French.
Methods
The research was conducted according to the proposals of Greenhalgh [44,45] and the search strategies of the Cochrane Oral Health Group, the Cochrane Neonatal Group, the Cochrane Pregnancy Group and the Cochrane Childbirth Group were applied. As a recent paper pointed out the truncation of orthodontic* as suggested by the Cochrane Oral Health group to be entailing the implicit exclusion of relevant articles [46] other search strategies were used in addition to those of the Cochrane groups: (((child* OR infant* OR (low birthweight) OR neonate* OR premature* OR preterm*) AND (alveol* OR gum* OR palate* OR maxill* OR orthodon* OR groov*)) NOT (syndrom* OR cleft* OR cancer* OR carcino* OR fract* OR traum* OR surg* OR infect* OR occlusion* OR malocclusion* OR laser* OR (orthodontic treatment) OR caries OR lung OR cell OR cancer [sb] OR space [sb] OR cam [sb] OR tox [sb] OR history [sb] OR aids [sb] OR letter [pt])) Field: Title/Abstract, Limits: All Child: 0–18 years, Human.
Electronic literature research comprised the following medical databases: AMED, BIOSIS, CareLit, Cochrane Library, Current Contents, EMBASE, KindHeilk, Oldmedline, Pubmed, Web of Science. Additionally, 'hand search' was performed in text books and encyclopedias relevant to the subject, and in the following journals, including supplements and abstract bands: Clinics in Perinatology, vol. 21–30 (1994–2003); Der Kinderarzt vol. 16–21 (1985–1990) and vol. 25–31 (1994–200); Der Kinder- und Jugendarzt vol. 32–33 (2001–2002); Kinderärztliche Praxis vol. 53–61 (1985–1993) and 68–73 (1997–2002); Pediatric Clinics vol. 32–50 (1985–2003); Pediatric Research vol. 19–33:1 (1985–1993), 33:3–52 (1995–2002) and Pediatrics vol. 75–110 (1985–2002). Retrieved publications were checked for references and, where appropiate, these publications found in the bibliographies were considered in the review.
Selection criteria
Sources in English, German and French were included from 1900 to 1/2004: non-metric visual findings, metric studies with intraoral measurements if no absolute numerical data were given but only visual findings were expressed in relative terms, and metric studies on plaster casts made from impressions of the palate. Only data provided in the reports were considered, and no attempt was made to contact the authors for missing or 'raw' data, because our research reached back to the beginning of the last century so that it would not have been possible to contact all authors.
As the authors of the review had been strongly involved in the subject matter, it was not possible to 'blind' them for the studies. Titles and/or abstracts of all citations were screened by one author (AH). In cases of doubt, the inclusion or exclusion of studies was discussed, and consensus was reached, among all authors. The full text of all relevant studies was evaluated. Exclusion criteria and affected studies are listed in Table 1 (see Additional file 1).
Results
The electronic search strategy resulted in 141 articles, six abstracts, four conference papers, eight letters, six dissertations, and two masters theses. By hand search, eight abstracts, six bookchapters, and one encyclopedia were identified. Twenty-eight studies were excluded (Table 1, see Additional file 1) and one hundred fifty-five articles remained for final analysis. Among these, one hundred nineteen studies assessed morphometrically the development of the palate.
The first identified study was published in 1934 [47]. Looking at the different publication years and the different research questions, the articles can be divided into two parts: morphology of the preterm palate and morphology of the term newborn palate. The latter research topic was of great interest before 1960, whereas research on the preterm palate had its peak in 1985.
A further pattern to classify the research is the general methodology used for morphological assessments. Two different approaches could be identified for both groups, term and preterm infants: visual descriptions and metrical descriptions of the palatal configuration. Therefore, the review presented here follows the given patterns of past research and is divided into two parts.
Part 1 summarizes the applied descriptions of the palate in the term newborn. Without knowledge of the term infant's normal oral structures, it would be impossible to recognize abnormalities in the preterm infant's palate. The review of papers and bookchapters using visual descriptions of the term infant's palate – which are important for a general overview for the clinician – is followed by a presentation of metrical studies, which are necessary to validate clinical impressions which are the major source for measurements. The analysis of the studies is therefore ordered as follows.
• Visual description of the palatal configuration of the term newborn
- Palatal configuration with respect to gender and race
• Metric description of the palatal configuration of fullterm infants
- Palatal configuration with respect to gender and race
- Palatal configuration with respect to cranial index
- Palatal configuration with respect to mode of delivery
- Palatal configuration with respect to nasal deformities
Visual description of the palatal configuration of the term newborn
In the newborn, the jaw already displays the palatine rugae present in the adult as well as the frenulum and the incisive papilla. In most cases the frenulum, which is located between the lip and the incisive papilla, recedes. If it fails to do so, it is known as a persisting tectolabial frenulum, which may later give rise to a midline diastema [48].
In addition to the structures present in the adult, however, the maxilla of the newborn is characterized by a special feature: the groove system (Figure 2). This separates clearly visible maxillofacial regions and valla from one another and, like the frenula, is subject to a non-uniform nomenclature [49]. According to an investigation on palatal casts of 500 newborn fullterm children, the maxillary alveolar arch is marked along its whole length by the dental groove which divides it into two parts, a lateral labiobuccal and a medial lingual portion; it is through the former of these that the teeth eventually erupt [50] (Figure 2). The gum pad is divided into ten segments [51] which correspond to the developing tooth germs [50]. The central incisor and canine segments are approximately equal in size and are well marked; they are separated from the smaller lateral incisor segment, which is indistinct and sometimes lies lingual to them, by two shallow grooves. The lateral sulcus runs anteriorly from the lingual to the labial aspects and sometimes extends to a lateral frenum, this sulcus is the anterior margin of the first deciduous molar segments, which are the largest [51]. The second molar segment is more difficult to recognize. Merging with the dental groove, it can be made out lying somewhat lingual to the first molar segment. The gum is solid and firm throughout. In the distal part of the maxilla, the pseudoalveolar ridge can be recognized, a transient structure, which disappears in the first months of life (Figure 2).
Figure 2 The alveolar portion of the upper gum pad is separated from the palate by a groove. The alveolar portion itself is again divided into buccal and lingual portions which are also separated by grooves. The former is the larger, participates in the formation of the sheath and socket of the teeth, and is further divided by transverse grooves or sulci into segments corresponding to the developing tooth germs. For nomenclature of palatal structures, see Table 2 (Additional file 2). Interestingly, the 'Terminologia Anatomica' contains for discription of palatal structures only the following terms: Frenulum labii superioris (Frenulum of upper lip); Palatum (Palate); Palatum durum (Hard palate); Palatum molle, Velum palatinum (Soft palate); Raphe palati (Palatine raphe); Plicae palatinae transversae, Rugae palatinae (Transverse palatine folds; Palatine rugae); Papilla incisiva (Incisive papilla).
What is of greatest importance within the framework of the present review is – as discussed in Part 3: 'consequences of intubation' – the most palatally located vallum (synonyms: tectal vallum, tectal ridge, lateral palatine ridge, lateral alveolar ridge, lateral palatine prominences or lateral palatine processes; see Figure 2, Table 2 (see Additional file 2)), which is a normal structure in the neonate and does not have an osseous but rather a connective-tissue base [52]. Hanson et al. reported after examining three deceased and 260 normal infants [52]: 'During early fetal life the lateral palatine ridges are composed of loose mesenchymal tissue with collagenous fibers embedded in lightly PAS-positive matrix. Alcian blue staining confirms the presence of acis mucopoloysaccharide. As development progresses the connective tissue become more dense, and the ridges appear less prominent in relation to the adjacent structures. As a result of the smoothing of the palatal vault and continued growth of the alveolar ridge, the lateral palatine ridges are less prominent in the normal full-term infant than in earlier stages.'
Although there are obvious growth changes, the gum pad shows similar features at six months of age [25] (Table 3, see Additional file 3). Probably due to tongue thrust into the palatine vault [53], there is then a marked flattening of the lateral palatine ridges in the second year of life [52]. In the vast majority of the normal children (48 out of 56) the lateral palatine ridges are no longer apparent at the age of five years [52]. The configuration of the palate is then similar to that observed in adults.
Klemke [54] reported in his study on 200 newborns various kinds of upper jaws: a nearly semicircular form, a shape with a flattened anterior part and a nearly eleptic arch (percentages not given) (Table 3, see Additional file 3). In accordance, Neumann [55] described individual variations in palatal shape of 200 newborns, the majority of children, however, having horse-shoe or u-shaped palates (no percentages given) (Table 3, see Additional file 3). Approximately 1/3 of her probands presented a parable shape. Ott [56] diagnosed characteristic changes with respect to palatal shape in the course of time: up to the age of twelve months the majority of jaws had a semicircular anterior form with convergent sides, from 16 to 24 months parallel sides, and from 28 to 32 months divergent sides. She interpreted those changes in connection with the tooth eruption. The reliability of the method was not given in all three dissertations [54-56] (Table 3, see Additional file 3).
All elements of the bony palate are present in the fullterm neonate. The median palatine suture is a firm, fibrous articulation without fusion. The transverse suture between the palatine process of the maxilla and the intermaxillary bone is usually open and closes during the first year [57]. The palate during the first year of life is relatively broad and flat [58].
Epstein's Pearls, currently called palatal cysts [59] are remnants of epithelial tissue trapped during the palatal fusion. Their general incidence has been reported to be around 65% in full term newborns [60,61]. Bohn's nodules are remnants of mucuous gland tissue found on the buccal or lingual aspects of the dental ridges, dental lamina cysts (glands of Serres) are found along the crest of the alveolar ridges, both together are currently called alveolar cysts, and have also been referred to as 'gingival cysts' or 'inclusion cysts' [59]. An incidence of 36% of maxillary alveolar cysts in 1 – 5 days old full term newborns is reported [60]. The clinical description of palatal and alveolar cysts varies in color from white, to gray to yellow nodules, in size from a pinhead to 3 mm, and in numbers from 1 to 6 [59].
Palatal configuration with respect to gender and race
According to Dittrich [62] in newborn infants of both sexes the predominant form of the upper jaw is that with semicircular anterior parts and converging sides (male: 62%, female: 66%), followed by that with parallel sides (male: 18%, female: 2%). In contrast, Oelschlaegel [63] found a significant higher percentage of girls (62%) with a semicircular anterior part of the palate than boys (51%), and observed in boys among all possible forms of the side the parallel form to be the most frequent. In neither of the studies the reliability of the method was given, nor was mentioned explicitely that term infants had been examined. Leighton and Seshadri [64] found in a sample of 34 Caucasian full term infants at birth in only 14.7% midline notching of the upper gum pad compared to 34 matched Afro-Carribean infants, who had in 67.6% of the cases midline notching. The distribution of sexes in the sample was not given.
Huddart and Graf [65] also revealed differences between English, Italian and Swiss babies, affecting principally the anterior part of the upper gum pad and the contour of the palate. In a study with 500 normal full term newborns (82% blacks, 18% whites) from in the majority economically disadvantaged mothers aged less than 25 years in an urban setting (none of the babies had been admitted to the intensive care unit) a total of 21% alveolar notches was found. Those notches were significantly more common in blacks, with an odds ratio of 2.7 (a definition of the notches and their grades of severity was not given) [66]. There is speculation that notching is associated with a midline diastema in the primary and permanent dentition [61,67].
Friend et al. [66] found in 58% of the above mentioned sample cysts within the median raphe or the hard palate (Epstein pearls) or palatal cysts, which were defined by the authors as Bohn's nodules, i.e. whitish nodules at the junction of the hard and soft palate adjacent to the midpalate raphe (n.b. the different terminology in comparison to Donley et al. [59]). Most of the lesion in series were found at the hard-soft palate juncture. Midpalatal cysts were 2.5 times more likely to occur in white newborns (75%) than in blacks (55%). With respect to mentioned palatal structures, no gender difference was found.
Monteleone and McLellan [68] described results similar to those of Friend et al. [66]. The former found palatal cysts in 85% of the white children in contrast to 79% of the black children. With respect to palatal and alveolar cysts, Donley and Nelson [59] did not find significant differences between a cohort of caucasian children and a group of non-caucasian infants established from black, Latino and Indian children. Nor did they find gender differences.
Friend et al. [66] found no palatal cyst in the premaxillary region, all were posterior to the incisive foramen, which might be explained by the fact that the premaxilla is the first portion of the palate to fuse, if the pathogenesis of these lesions depends on entrapment of epithelium during fusion of the palatal shelves [69]. Alveolar cysts (grayish-white nodules along the crest of the alveolar mucosa or, less commonly on the lingual or facial borders) also were more likely in whites (26%) than in black children (11%, odds ratio 3.3, no distinction between upper and lower jaw was given).
Jorgenson et al. [61] found a higher total incidence in both races, but also diagnosed more alveolar cysts in whites (53%) than in blacks (40%). Although palatal and alveolar cysts are similar clinically and histologically, the former were more common, a discrepency which might be explained by the histopatholic presence of alveolar cysts in stillborns with lacking clinical manifestation [60].
Alveolar lymphangioma (blue, domed, fluid filled lesions on the alveolar ridges of either arcade, which occur typically bilaterally) was only found in black children, with teenage mothers at enhanced risk of having a child with this condition [66]. The latter authors found in none of the mentioned parameters a predeliction in gender, nor did Donley and Nelson [59].
Metric description of the palatal configuration of fullterm infants
In order to validate clinical impressions, there is a need for measuring palates. Manufactureres would benefit from metric information to design save products [70].
Valid, plaster cast-based metric descriptions of the palatal configuration of healthy, term children around birth and in the first years of live are, however, rare (Table 3, see Additional file 3; Part 2: Table 4, see Additional file 4 of Part 2). Additionally, the following difficulties occur: in only four studies is named explicitely that term infants have been examined [25,50,67,71]; in two further studies, weight and/or maturity of the included children were given, enabling the authors of the review to recalculate the data for term infants; only eight studies [25,51,54,55,58,59,67,71] provided data on the weight and/or body length of the probands. Thus, the comparability of the results is limited.
The form of the upper jaw can considerably vary [54]. By recalculation of original data for term infants a correlation between maximum palatal width and weight, length of body, and biggest head circumference could be found (p < .05) [54].
In contrast, Leighton observed a low correlation (r = 0.4) between the dental arch width of neonates and their birthweight [67]. He moreover detected in a comparison of monozygotic twins, dizygotic same-sex twins and dizygotic different-sex twins that the differences in palatal width were twice as large in the last group as in the first. The author interpreted this as an indication that palatal width is genetically determined. Although genetic influence was clearly an important factor in determining gum pad morphology, there was only a weak correlation of size between the contained deciduous tooth crowns and the upper gum pad in the newborn [67], which may be due to the thick pad of fibrous tissue overlaying the developing tooth germ [25]. At the age of six months, however, the sum of the upper tooth diameters correlates significantly with maximum palatal width and postgingival width, as does weight. The size of the alveolar process is more linked to tooth size, whereas the total size of the gum pad, and the palate in particular, is more closely related to bodyweight. Sucking habits show a small but significant correlation with the stated palatal parameters, too, suggesting that a sucking habit is associated with narrowing of the palate, without its area or anterior length being altered. Age showed no significant correlation. In addition, the twin-based research revealed a significant hereditary influence on palatal width at birth [64]. A significant heriditary influence on the postgingival width was determined in a comparison of approximately 6-month-old identical twins versus fraternal, dissimilar-sex twins, but not versus fraternal, same-sex twins [25].
In six infants a pronounced transversal growth in the first six months of life was described, wheras a sagittal growth about zero was reported [72]. In contrast, in a study with 428 children, maximum palatal width and maximum palatal length grew at about the same rate in the first year, the width-length index remaining unchanged during the first year. Palatal height increased until 9 months, then remained quite constant up to the first year. The palatal dimensions varied widely during the first year, by about 7% for the width and length dimensions and by about 10 – 12% for maximum height (percentage variability means the coefficient of variation, i.e. the standard deviation divided by the means, times 100) [58]. A continuous increase in mean maximum palatal width of on average 7.45 mm from birth was measured (30.99 mm, value taken from [62]) to 32 months of age (38.44 mm), with the least changes occuring from 12 – 28 months [56]. The increase of mean length of the upper jaw was 9.18 mm (from 24.58 mm to 33.76 mm) (measurements from the beginning of the labial frenulum to a connecting line between the tubera).
Palatal configuration with respect to gender and race
The existence of ethnic differences could be important to those responsible for overseeing oral development of neonates in populations containing members of different ethnic groups.
Significant greater widths in the anterior part of the gum pad of 34 Afro-Carribean full term infants compared to 34 Caucasian full term subjects have been described [67]. No significant differences neither in the width of the gum pads distal to the lateral sulci nor in palatal height were detected. The differences in height only achieved statistical significance when expressed as the ratio of maximum width to palatal height. Unfortunately, the sex of the sample was not given so that the differences in size could have been wrongly attributed to races but could indeed have also occured due to gender differences: another study reveals the anterior parts of the upper arch of 7–12 years old girls to be smaller, but the posterior parts to be wider than those of the boys [73].
By recalculating the figures given by Neumann [55] no significant gender differences with respect to palatal width were found by the authors of the present review for spontaneously delivered term children aged 1–7 days, matched for birthweight and size (occipito-anterior vertex presentation exclusively). This is in contrast to the results given by the author herself, who found a significant sex difference for palatal width, but who did not distinguish between term and preterm children, included children up to three weeks of age, and did not consider the mode of presentation. Correlations between palatal width and size or birthweight could not be found, either, which is in accordance with the original results given by Neumann [55], and means that newborn children of the same size and birthweight can have differently dimensioned palates. The correlation coefficients for maximum palatal width and length with bodyweight and total body length in 100 male newborns were shown to be of low order (between 0.37 and 0.56), as the palatal dimensions are poorly correlated with each other and with other body dimensions. The palatal dimensions in the male are on average larger than in the female, corresponding to the larger mean size of male newborns [58].
Dittrich [62] and Oelschlägel [63] found significantly wider palates in boys compared to girls, the latter also significantly deeper palates in boys. Oelschlägel [63] did not find any gender differences for palatal length. In contrast, Dittrich measured significantly longer palates in boys [62]. This is in accordance with Hall et al. [48], who found the gender-related difference in palatal length to be accentuated with increasing age (reference values for palatal length, height and width are not given until the age of 5).
Palatal configuration with respect to cranial index
No correlation between cranial index (biparietal diameter in percentage of frontooccipital diameter) and palatal index (breadth in percentage of length) was found in 515 boys and 455 girls [63].
Palatal configuration with respect to mode of delivery
Every baby is subject to a certain amount of pressure during parturation, with head adaptations such as parietal bone and facial molding. Any molding of the face occurs across the maxilla, because the bimalar span is the widest part of the face. This molding compresses and deforms the soft maxilla, resulting in possible elevation of the arch of the palate [74].
No significant differences in palatal lengths, depths and widths dimensions between infants with spontaneous vertex presentation (n = 89, age 8 days, > 3.4 kg) and children with high or low forceps delivery (n = 10, age 8 days, > 3.4 kg) were described [51]. Klemke [54] and Hofbauer [75] did not detect an influence of mode of delivery on the jaws, either. Data for elective caesarian section and spontaneous face presentation was to small to draw any conclusions [51,55].
Palatal configuration with respect to nasal deformities
Kent et al. [71] tested the hypotheses made by Gray [76] that pressures in the maxilla during birth may cause elevation in one side of the palate and thus asymmetry of the hard palate which in turn could distort the vomer and septal cartilage. The former authors found no evidence of palatal asymmetry (Table 3, see Additional file 3) in 14 out of 500 children compared to 14 controls within three days of birth. The method of measurement was, however, quite coarse and the reliability of the method not given.
In contrast, at ages 3 – 6 years (n = 145) [74], 5 – 6 years (n = 145) and again in children 'aged about 8 years' (n = 90) [77] statistically significant more often palatal asymmetries were found in children whose nasal septae were not in the midline at birth. In both studies, height of the palate was related neither to the evenness respectively unevenness of the palate nor to the type of septal deformity [74,77]. However, palatal asymmetry of width and height was present statistically significant most frequently in septae kinked to one side, less by septae deviated to both sides and least by straight septae. The method of measurement and the error of the method were not given in either of the two studies.
List of abbreviations
[PT] preterm infant, [BW] birthweight, [LBW] low birthweight, [NBW] normal birthweight, [VLBW] very low birthweight, [NBW] normal birthweight, [GA] gestational age, [GW] gestational weeks, [NS] not significant
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
AH designed the study, searched the databases, extracted the data, analyzed the results and wrote the manuscript. HR helped with study design, analysis and provided critical input in neonatal associated issues and revised the manuscript. UE and EH formulated the research question, helped with study design, analysis and in revising the manuscript. All authors read and approved the final manuscript.
Supplementary Material
Additional File 1
Table 1. Excluded studies and reasons for exclusion.
Click here for file
Additional File 3
Table 3. Studies on plaster casts.
Click here for file
Additional File 2
Table 2. Nomenclature of palatal structures as presented in Figure 2.
Click here for file
Acknowledgements
We thank Fiona Lawson for the English language revision.
==== Refs
Usher R Avery G The special problems of the premature infant Neonatology: Pathophysiology and Management of the Newborn 1981 2 Philadephia: JB Lippincott
Seow WK Masel JP Weir C Tudehope DI Mineral deficiency in the pathogenesis of enamel hypoplasia in prematurely born, very low birthweight children Pediatr Dent 1989 11 297 302 2639324
Garbe W Das Frühchen – Buch 1997 Stuttgart: Thieme
Fearne JM Brook AH Small primary tooth-crown size in low birthweight children Early Hum Dev 1993 33 81 90 8055779 10.1016/0378-3782(93)90203-7
Evaldson G Lagrelius A Winiarski J Premature rupture of the membranes Acta Obstet Gynecol Scand 1980 59 385 393 7446003
Klinisches Wörterbuch Psychrembel 1990 256 Berlin: W de Gruyter
Copper RL Goldenberg RL Das A Elder N Swain M Norman G Ramsey R Cotroneo P Collins BA Johnson F Jones P Meier AM The preterm prediction study: maternal stress is associated with spontaneous preterm birth at less than thirty-five weeks' gestation. National Institute of Child Health and Human Development Maternal-Fetal Medicine Units Network Am J Obstet Gynecol 1996 175 1286 1292 8942502 10.1016/S0002-9378(96)70042-X
Walker BR McConnachie A Noon JP Webb DJ Watt GC Contribution of parental blood pressures to association between low birth weight and adult high blood pressure: cross sectional study BMJ 1998 316 834 837 9549456
Hillier SL Nugent RP Eschenbach DA Krohn MA Gibbs RS Martin DH Cotch MF Edelman R Pastorek JGn Rao AV Association between bacterial vaginosis and preterm delivery of a low-birth-weight infant. The Vaginal Infections and Prematurity Study Group 1995
Boyd DL Katz V Ferktik GS Offenbacher S Relationship between periodontal disease status and birth outcome [abstract] J Dent Res 1994 73 s394
Collins G Windley HW Arnold RR Offenbacher S Effects of a Porphyromonas gingivalis infection on inflammatory mediator response and pregnancy outcome in hamsters Infect Immun 1994 62 4356 4361 7927695
Windley HW IIICollins JG Arnold RR Offenbacher S Molecular mechanisms of P. gingivalis-induced pregnacy abnormalties in hamster model [abstract] J Dent Res 1994 189 s702
Offenbacher S Katz V Fertik G Collins J Boyd D Maynor G McKaig R Beck J Periodontal infection as a possible risk factor for preterm low birth weight J Periodontol 1996 1103 1113 8910829
Lopez NJ Smith PC Gutierrez J Higher risk of preterm birth and low birth weight in women with periodontal disease J Dent Res 2002 81 58 63 11820369
Davenport ES Williams CECS Sterne JAC Murad S Sivapathasundram V Curtis MA Maternal periodontal disease and preterm low birthweight: case-control study J Dent Res 2002 81 313 8 12097443
Kovar IZ Mayne PD Robbe I Hypophosphataemic rickets in the preterm infant; hypocalcaemia after calcium and phosphorus supplementation Arch Dis Child 1983 58 629 631 6614979
Seino Y Ishii T Shimotsuji T Ishida M Yabuuchi H Plasma active vitamin D concentration in low birthweight infants with rickets and its response to vitamin D treatment Arch Dis Child 1981 56 628 632 6973954
Harris EF Barcroft BD Haydar S Haydar B Delayed tooth formation in low birthweight African-American children Pediatr Dent 1993 15 30 35 8233989
Seow WK Brown JP Tudehope DI O'Callaghan M Developmental defects in the primary dentition of low birth-weight infants: adverse effects of laryngoscopy and prolonged endotracheal intubation Pediatr Dent 1984 6 28 31 6592545
Burns Y Rogers Y Neil M Development of oral function in pre-term infants Physiother Pract 1987 3 168 178
Rickards AL Ford GW Kitchen WH Doyle LW Lissenden JV Keith CG Extremely-low-birthweight infants: neurological, psychological, growth and health status beyond five years of age Med J Aust 1987 147 476 481 2446110
Eight-year outcome in infants with birth weight of 500 to 999 grams: continuing regional study of 1979 and 1980 births. Victorian Infant Collaborative Study Group J Pediatr 1991 118 761 767 2019933
Escobar GJ Littenberg B Petitti DB Outcome among surviving very low birthweight infants: a meta-analysis Arch Dis Child 1991 66 204 211 1825773
Fadavi S Adeni S Dziedzic K Punwani I Vidyasagar D The oral effects of orotracheal intubation in prematurely born preschoolers ASDC J Dent Child 1992 59 420 424 1491080
Leighton BC A preliminary study of the morphology of the upper gum pad at the age of 6 months Swed Dent J 1982 115 122
McCormick MC The contribution of low birth weight to infant mortality and childhood morbidity N Engl J Med 1985 312 82 90 3880598
Kim Seow W Tudehope DI Brown JP O'Callaghan M Effect of neonatal laryngoscopy and endotracheal intubation on palatal symmetry in two- to five-year old children Pediatr Dent 1985 7 30 36 3857557
Morris KM Seow WK Burns YR Palatal measurements of prematurely born, very low birth weight infants: comparison of three methods Am J Orthod Dentofacial Orthop 1993 103 368 373 8480703
Viscardi RM Romberg E Abrams RG Delayed primary tooth eruption in premature infants: relationship to neonatal factors Pediatr Dent 1994 16 23 28 8015938
Davidson S Schrayer A Wielunsky E Krikler R Lilos P Reisner SH Energy intake, growth, and development in ventilated very-low-birth-weight infants with and without bronchopulmonary dysplasia Am J Dis Child 1990 144 553 559 2330922
Hack M Breslau N Weissman B Aram D Klein N Borawski E Effect of very low birth weight and subnormal head size on cognitive abilities at school age N Engl J Med 1991 325 231 237 2057024
Vaucher YE Harker L Merritt TA Hallman M Gist K Bejar R Heldt GP Edwards D Pohjavuori M Outcome at twelve months of adjusted age in very low birth weight infants with lung immaturity: a randomized, placebo-controlled trial of human surfactant J Pediatr 1993 122 126 132 8419599
Bosma JF Heriza C Examination of the mouth and pharynx of the infant Proceedings: The comprehensive management of infants at risk for CNS deficits 1974 Department of Medical Allied Health Professions, University of North Carolina, Chapel Hill
Palmer B The influence of breast feeding on the development of the oral cavity J Hum Lactat 1998 14 93 98
Pirttiniemi P Grön M Alvesalo L Heikkinen T Osborne R Relationship of difficult forceps delivery to dental arches and occlusion Pediatr Dent 1994 16 289 293 7937262
Paulsson L Bondemark L Soderfeldt B A systematic review of the consequences of premature birth on palatal morphology, dental occlusion, tooth-crown dimensions, and tooth maturity and eruption Angle Orthod 2004 74 269 279 15132456
Jones DR Meta-analysis of observational epidemiological studies: a review J R Soc Med 1992 85 165 168 1556722
Derks A Katsaros C Frencken JE van't Hof MA Kuijpers-Jagtman AM Caries-inhibiting effect of preventive measures during orthodontic treatment with fixed appliances. A systematic review Caries Res 2004 38 413 420 15316184 10.1159/000079621
O'Neill J Little evidence exists about optimal caries-prevention strategies during orthodontic treatment Evid Based Dent 2004 5 97 15608709 10.1038/sj.ebd.6400300
Lagravere MO Major PW Flores-Mir C Long-term dental arch changes after rapid maxillary expansion treatment: a systematic review Angle Orthod 2005 75 155 161 15825776
Lagravere MO Major PW Flores-Mir C Skeletal and dental changes with fixed slow maxillary expansion treatment: a systematic review J Am Dent Assoc 2005 136 194 199 15782523
Hunt O Burden D Hepper P Johnston C The psychosocial effects of cleft lip and palate: a systematic review Eur J Orthod 2005 27 274 285 15947228 10.1093/ejo/cji004
Benson PE Shah AA Millett DT Dyer F Parkin N Vine RS Fluorides, orthodontics and demineralization: a systematic review J Orthod 2005 32 102 114 15994984 10.1179/146531205225021033
Greenhalgh T How to read a paper. The medline database BMJ 1997 315 180 183 9251552
Greenhalgh T How to read a paper. Papers that summerize other papers [systemtic reviews and meta-analyses] BMJ 1997 315 672 675 9310574
Stamm T Hohoff A Nonlinear behavior of search strategies for identifying relevant orthodontic articles Angle Orthod 2004 74 316 318 15264640
Ashley-Montagu MF The form and dimension of the palate in the newborn Int J Orthod Dent Child 1934 20 694 827 10.1016/S0097-0522(34)90177-5
Hall J Froster-Iskenius U Allanson J Handbook of normal physical measurements 1989 Oxford, New York, Toronto: Oxford University Press
Miethke R Zur intrauterinen Entwicklung der Kiefer und Lippen bei menschlichen Feten von der 17. bis zur 42. Woche Habilitationsschriften der Zahn- Mund and Kieferheilkunde 1979 Berlin: Quintessenz Verlags-GmbH
Clinch L Variations in the mutual relationships of the maxillary and mandibular gum pads in the newborn child Int J Orthod 1934 20 359 374
Sillman JH Relationship of maxillary and mandibular gum pads in the newborn infant Am J Orthodont Oral Surg 1938 24 409 424 10.1016/S0096-6347(38)90295-9
Hanson JW Smith DW Cohen MMJ Prominent lateral palatine ridges: developmental and clinical relevance J Pediatr 1976 89 54 58 932903
Behrstock B Ramos A Kaufman N Does prolonged oral intubation contribute to medical hypertrophy of the lateral palatine ridges and possibly to iatrogenic cleft palate? J Pediatr 1977 91 171 874659
Klemke B Über Kieferform und Bisslage beim Neugeborenen Med Diss Bonn 1939
Neumann M Kieferbezügliche Untersuchungen und Messungen an Neugeborenen Med Diss Kiel 1953
Ott H Beitrag zur normalen Entwicklung des Milchgebisses von der Geburt bis zu ca. drei Jahren (anhand von ca. 463 Untersuchungsfällen) Med Diss Leipzig 1961
Erenberg A Nowak AJ Appliance for stabilizing orogastric and orotracheal tubes in infants Crit Care Med 1984 12 669 671 6430649
Bakwin H Bakwin R Form and dimension of the palate during the first year of life Int J Orthod 1936 22 1018 1024
Donley CL Nelson LP Comparison of palatal and alveolar cysts of the newborn in premature and full-term infants Pediatr Dent 2000 22 321 324 10969441
Cataldo E Berkman MD Cysts of the oral mucosa in newborns Am J Dis Child 1968 116 44 48 5657354
Jorgenson RJ Shapiro SD Salinas CF Levin LS Intraoral findings and anomalies in neonates Pediatrics 1982 69 577 582 7079011
Dittrich I Untersuchungen über Form and Gröβe des Neugeborenen – Oberkiefers anhand von 1000 eigens gewonnenen Modellen Med Diss Leipzig 1959
Oelschlägl S Gestaltuntersuchungen an Oberkiefermodellen von 1000 Neugeborenen unter Berücksichtigung des Geschlechtsunterschiedes und der Beziehung zum Kopfindex Med Diss Leipzig 1954
Leighton B Seshadri B A comparison of the gum pads of Afro-Caribbean and Caucasian British subjects at birth Br J Orthod 1990 17 215 221 2207052
Huddart A Graf B Kriens O The maxillary arch dimensions of English, Swiss and Italian normal children at birth and 4 months. Are there national differences? What is cleft lip and palate? A multidisciplinary update; proceedings of an advanced workshop 1987 Stuttgart: Thieme
Friend GW Harris EF Mincer HH Fong TL Carruth KR Oral anomalies in the neonate, by race and gender, in an urban setting Pediatr Dent 1990 12 157 161 2077490
Leighton BC Morphologische Variationen der Alveolarbögen beim Neugeborenen Fortschr Kieferorthop 1976 37 8 14 10.1007/BF02169375
Monteleone L Mclellan MS Epstein's pearls (Bohn's nodules) of the palate J Oral Surg Anesth Hosp Dent Serv 1964 22 301 304 14155470
Fromm A Epstein's pearls, Bohn's nodules and inclusion-cysts of the oral cavity J Dent Child 1967 34 275 287 5342399
Nowak AJ Casamassimo PS Oral opening and other selected facial dimensions of children 6 weeks to 36 months of age J Oral Maxillofac Surg 1994 52 845 847 8040739
Kent SE Rock WP Nahl SS Brain DJ The relationship of nasal septal deformity and palatal symmetry in neonates J Laryngol Otol 1991 105 424 427 2072008
Deffez JP Plante P The maxilla of infants Rev Stomatol Chir Maxillofac 1976 77 403 408 1066765
Cohen JT Growth and development of the dental arches in children J Am Dent Assoc 1940 27 1250 1260
Gray LP Deviated nasal septum. Incidence and etiology Ann Otol Rhinol Laryngol Suppl 1978 87 3 20 99070
Hofbauer K Haben Geburtslage und Geburtstrauma Einfluss auf die Entstehung von Regelwidrigkeiten des Gebisses, insbesondere der Bisslage? Dtsch Zahn- Mund- und Kieferheilkunde 1943 19 356 368
Gray LP Dillon PI Brogan WF Henry PJ The development of septal and dental deformity from birth Angle Orthod 1982 52 265 278 6961828
Gray LP The development and significance of septal and dental deformity from birth to eight years Int J Pediatr Otorhinolaryngol 1983 6 265 277 6676334
Alkan L Gewisse Formen des harten Gaumens und ihre Entstehung Arch Laryng Rhin 1990 10 441 463
Brawley RE Sedwick HJ Studies concerning the oral cavity and saliva. I. Palate (2). Palatal measurements Am J Orthod Oral Surg 1939 25 1062 1068 10.1016/S0096-6347(39)90202-4
Channing W Wissler C Comparative measurements of the hard palate in normal and feeble minded individuals. A preliminary report Am J Insan 1905 61 687 697
Delaire J Considérations sur l'accroissement du pré-maxillaire chez l'homme Rev Stomatol Chir Maxillofac 1974 75 951 970 4533145
Denzer BS The size of the infantile palate Am J Dis Child 1921 22 471 476
Denzer BS The size of the infantile palate Int J Orthodont Oral Surg Radiography 1922 8 510 515
Franke G Über Wachstum und Verbildungen des Kiefers und der Nasenscheidewand aufgrund vergleichender Kiefermessungen und experimenteller Untersuchungen über Knochenwachstum Zeitschr f Laryng Rhin Otol 1922 10 187 201
Freiband B Growth of the palate in the human fetus J Dent Res 1937 16 103 122
Friel S The development of ideal occlusion of the gum pads and the teeth Am J Orthodont 1937 40 196 227 10.1016/0002-9416(54)90145-1
Heckmann U Brune K Grabowsky R Über das Breitenwachstum der Kiefer Dtsch Stomat 1969 19 759 766
Hoffer O Entwicklungsbezügliche Feststellungen am Kauorgan von Föten and Neugeborenen Fortschritte der Kieferorthopädie 1956 17 166 176 10.1007/BF02169743
Huddart AG Clarke J Thacker T The application of computers to the study of maxillary arch dimensions Brit Dent J 1971 130 397 404 5280980 10.1038/sj.bdj.4802671
Huddart AG MacCauley FJ Davis MEH Maxillary arch dimensions in normal and unilateral cleft palate subjects Cleft Palate J 1969 6 471 487 5260755
Katz J Über die Gaumenform beim Neugeborenen Med Diss Frankfurt 1923
Kopra DE Davis EL Prevalence of oral defects among neonatally intubated 3- to 5- and 7- to 10-year old children Pediatr Dent 1991 13 349 355 1843991
Lang J Baumeister R Postnatale Entwicklung der Gaumenbreite und -höhe und die Foramina palatina Anat Anz Jena 1984 155 151 167
Lavelle CLB The shape of the dental arch Am J Orthod 1975 67 176 184 1054217 10.1016/0002-9416(75)90068-8
Lebret L Growth changes of the palate J Dent Res 1962 41 1391 1404
Lebret L Der menschliche Gaumen: Sein Wachstum, die auf ihn bezogene Wanderung der Seitenzähne, seine Expansion bei Anwendung zweier verschiedener orthodontischer Behandlungsmethoden Fortschr Kieferorthop 1966 27 121 140 5233533 10.1007/BF02163901
Macey-Dare LV Moles DR Evans RD Nixon F Long-term effect of neonatal endotracheal intubation on palatal form and symmetry in 8–11-year-old children Eur J Orthod 1999 21 703 710 10665200 10.1093/ejo/21.6.703
Melsen B Palatal growth studies on human autosy material. A histologic microradiographic study Am J Orthod 1975 68 42 54 1056143 10.1016/0002-9416(75)90158-X
Melsen B Melsen F The postnatal development of the palatomaxillary region studied on human autopsy material Am J Orthod 1982 82 329 432 6961805 10.1016/0002-9416(82)90467-5
Mithke RR Die Oberkieferformen menschlicher Feten vom 5. bis zum 10. Lebensmonat ZWR 1979 88 948 951 294007
Miotti F Miotti A Drusini A Sagittal dimensions of the anterior cranial base, maxilla and mandibular body in newborn male infants Eur J Orthod 1983 5 245 248 6578048
Peyton WT The dimension and growth of the palate in the normal infant and in the infant with gross maldevelopment of the upper lip and palate Arch Surg 1931 22 704 706
Poswillo D Observations of fetal posture and causal mechanisms of congenital deformity of palate, mandible and limbs J Dent Res 1966 45 584 596
Redman RS Shapiro BL Gorlin RJ Measurement of normal and reportedly malformed palatal vaults. II. Normal juvenile measurements J Dent Res 1966 45 266 269 5220218
Schulze C Die normale and abnorme Entwicklung des Gebisses Die Umbauvorgänge im Parodontium and Kiefergelenksbereich Morpho- bzw Pathogenese and Ätiologie der Dysgnathien 1993 2 No. 3 in Lehrbuch der Kieferorthopädie, Berin, Chicago, London, Sao Paulo, Tokio: Quintessenz Verlags-GmbH
Sillman JH Dimensional changes of the dental arches: Longitudinal study from birth to 25 years Am J Orthod 1964 50 824 842 10.1016/0002-9416(64)90040-5
Angelos GM Smith DR Jorgenson R Sweeney EA Oral complications associated with neonatal oral tracheal intubation: a critical review Pediatr Dent 1989 11 133 140 2668901
Kahl-Nieke B Einführung in die Kieferorthopädie 1995 München: Urban & Schwarzenberg
Ginoza G Cortez S Modanlou HD Prevention of palatal groove formation in premature neonates requiring intubation J Pediatr 1989 115 133 135 2738781
Howell S Assessment of palatal height in children Community Dent Oral Epidemiol 1981 9 44 47 6941877
Robke E Die Gaumenplatte nach Castillo Morales and ihre Wirkung auf die Gaumenentwicklung von Säuglingen mit Trisomie 21 im Vergleich zu gesunden Säuglingen Zahnmed Diss Münster 1998
|
16270908
|
PMC1308841
|
CC BY
|
2021-01-04 16:37:14
|
no
|
Head Face Med. 2005 Oct 28; 1:8
|
utf-8
|
Head Face Med
| 2,005 |
10.1186/1746-160X-1-8
|
oa_comm
|
==== Front
Health Qual Life OutcomesHealth and Quality of Life Outcomes1477-7525BioMed Central London 1477-7525-3-681627766210.1186/1477-7525-3-68ResearchEvaluating change in health-related quality of life in adult rhinitis: Responsiveness of the Rhinosinusitis Disability Index Chen Hubert [email protected] Patricia P [email protected] Stephen [email protected] Paul D [email protected] Department of Medicine, University of California, San Francisco (UCSF), CA, USA2 Cardiovascular Research Institute, UCSF, CA, USA3 Institute for Health Policy Studies, UCSF, CA, USA4 Department of Biostatistics and Epidemiology, UCSF, CA, USA2005 8 11 2005 3 68 68 24 6 2005 8 11 2005 Copyright © 2005 Chen et al; licensee BioMed Central Ltd.2005Chen et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The Rhinosinusitis Disability Index (RSDI) is a validated measure of health-related quality of life (HRQL) in rhinitis. Responsiveness of the RSDI to changes in health status over time has not been described.
Methods
We studied adults with a self-reported physician diagnosis of rhinitis identified through a national telephone survey. HRQL was assessed at baseline and at 24 months using the RSDI. Symptom severity, physical health status (SF-12 PCS), psychological mood (CES-D), and perceived control of symptoms were also assessed at the time of each interview. In addition, we ascertained specific health outcomes attributed to rhinitis, including days of restricted activity, job effectiveness, number of physician visits, and medication costs.
Results
Of 109 subjects interviewed at baseline, 69 (63%) were re-interviewed 24 months later. RSDI scores improved by = 0.5 standardized response mean in 13 (19%) subjects and worsened in 17 (25%). Change in the RSDI over time correlated with changes in symptom severity (r = 0.38, p = 0.001), physical health (r = -0.39, p = 0.001), mood (r = 0.37, p = 0.002) and perceived control of symptoms (r = -0.37, p = 0.01). In multivariate analyses adjusted for baseline health status, improvement in RSDI was associated with less restricted activity (p = 0.01), increased job effectiveness (p = 0.03), and decreased medication costs (p = 0.05), but was not associated with change in the number of physician visits from baseline (p = 0.45).
Conclusion
The RSDI is responsive to changes in health status and predicts rhinitis-specific health outcomes.
==== Body
Background
Rhinitis is a common chronic condition and can be a significant source of impairment. To measure the impact of rhinitis on health-related quality of life (HRQL), several disease-specific measures have been developed [1,2]. These instruments vary in length and content. Certain instruments, such as the 28-item Juniper Rhinoconjunctivitis Quality of Life Questionnaire (RQLQ), were originally intended to focus on persons with allergic disease [3]. Other instruments, such as the 16-item Sino-Nasal Outcome Test (SNOT-16), focus more on conditions characterized by chronic nasal obstruction, typically sinusitis [4].
The Rhinosinusitis Disability Index (RSDI), another measure of HRQL, has been used both in persons with allergic disease as well as in those with chronic nasal obstruction [5,6]. The instrument has 30 items comprising three domains: physical, functional, and emotional. Internal consistency and test-retest reliability of the instrument were initially described in 87 patients with a physician's diagnosis of rhinitis or rhinosinusitis [5]. The RSDI has subsequently been applied in larger populations across a spectrum of rhinologic diagnoses, ranging from allergic rhinitis to chronic sinusitis [6].
In a previous analysis, we evaluated the performance of the RSDI cross-sectionally in 109 adults with predominately allergic disease [7]. We found that the RSDI correlated in the expected ways with other measures of physical and mental health status, providing good evidence of construct validity. Furthermore, we found that psychosocial factors, such as the perception of one's own ability to deal with his or her nasal condition (termed 'perceived control'), can also play a large role in determining quality of life in rhinitis.
Although the discriminative properties of the RSDI have been studied in a cross-sectional fashion, its responsiveness to changes in health status has not been established. Responsiveness describes the ability of an instrument to capture meaningful changes in health over time. To demonstrate responsiveness, the same instrument must be administered at least twice over a given period of time. Responsiveness, however, differs from test-retest reliability. Test-retest reliability is measured by re-administering the same instrument within a short time interval during which the subject's health status remains unchanged. In contrast, responsiveness is typically demonstrated over longer intervals during which the subject's health status might reasonably have changed in some relevant way. Responsiveness has important implications for instruments used in clinical trials, where investigators are often trying to measure a clinically meaningful change before and after an intervention. In addition, it is also relevant in the study of disease progression or remission.
To determine the responsiveness of the RSDI, we studied change in HRQL in 69 adults with rhinitis using data collected over a 24-month period. Cross-sectional results of the 109 subjects studied at baseline have previously been reported [7]. Responsiveness of the RSDI was tested relative to other health status measures simultaneously administered to assess symptom severity, physical functioning, mood, and perceived control of symptoms. We then used the RSDI to evaluate the relationship between change in HRQL and other rhinitis-specific health outcomes.
Methods
Overview
We used data collected as part of a prospective, longitudinal cohort study of adults with asthma and rhinitis. We assessed change in HRQL and health status among 69 subjects with rhinitis alone (without concomitant asthma) who participated in two consecutive interview waves conducted approximately 24 months apart. We evaluated responsiveness of the RSDI relative to other generic and disease-specific health status measures simultaneously assessed. We also evaluated how change in disease-specific HRQL, measured by RSDI, was related to change in other rhinitis-specific health outcomes, including restricted activity, job effectiveness, physician visits, and medication costs.
Subject selection
Subject data were drawn from a larger prospective study of persons with asthma, rhinitis, or both. Details of recruitment and subsequent interview waves have been previously reported [8]. In brief, we enrolled English- and Spanish-speaking subjects aged 18 to 50 years from the Northern California region via random-digit dialing. Approval for the study of human subjects was obtained from the University of California, San Francisco Committee on Human Research.
We used data from interviews conducted in 2000–1 (baseline for this study) and 2002–3 (follow-up for this study). Subjects were included in our analysis if they reported a physician's diagnosis of allergic rhinitis, sinusitis, hay fever, or chronic post nasal drip, without concomitant asthma. One hundred nine subjects completed the rhinitis component of our health survey at baseline, of which 69 (63%) were re-interviewed at follow-up.
Health status measures
Rhinosinusitis Disability Index (RSDI)
To measure disease-specific HRQL in rhinitis, we used the RSDI, a 30-item questionnaire developed for use in persons with nasal or sinus disease [5]. Each item is rated on a 5-point Likert scale ranging from 'never' (scored as 0) to 'always' (scored as 4). The total score possible, calculated by summing the individual items, ranges from 0 to 120, with higher scores reflecting worse HRQL. The RSDI has 3 subscale domains: physical (11 items), functional (9 items), and emotional (10 items). The individual items comprising each of these domains are shown in Additional file: 1. Its reliability and validity have been demonstrated in patients with various rhinologic conditions [6]. In our analysis of subjects at baseline, the RSDI demonstrated high internal consistency with a Cronbach's alpha of 0.97 [7].
Rhinitis Symptom Score (RSS-4)
To measure symptom severity in rhinitis, we used four items drawn from a previously validated 31-item symptom scale for rhinoconjunctivitis and asthma developed by Wasserfallen et al [9]. The specific items included in our rhinitis questionnaire assessed four types of symptoms: 'sensation of fullness, congestion or blockage of the nose', 'sneezing', 'sinus headache or pain in face', and 'postnasal drip in back of throat'. Symptoms were rated on a 5-point Likert scale. Scores for the individual items were totaled yielding a possible range of 0 to 16, with higher score reflecting greater symptom severity. Using this 4-item symptom scale in our baseline analysis, we observed good internal consistency with a Cronbach alpha of 0.81.
The Short Form 12 (SF-12)
To measure general physical functioning we used the physical component summary (PCS) of the SF-12. The SF-12 consists of 12 items drawn from the widely used Short Form 36 (SF-36). The reliability, validity, and responsiveness of the SF-12 have been demonstrated in various disease states, and are comparable to those of the SF-36 [10]. SF-12 PCS scores range from 0 to 100, with a general population mean of 50 ± 10 (SD). Higher scores reflect better physical functioning.
Center for Epidemiologic Studies Depression scale (CES-D)
The CES-D consists of 20 items originally intended to assess depressive symptoms [11]. Further evidence, however, suggests that the CES-D also measures general psychological factors in addition to depression[12,13] For this reason, we refer to the CES-D as assessing 'psychological mood'. CES-D scores range from 0 to 60, with higher scores reflecting worse mood.
Perceived Control of Rhinitis Questionnaire (PCRQ)
Psychosocial factors such as an individual's perception of his or her ability to deal with illness, referred to as perceived control, have been demonstrated to be an important correlate of health status, both in rhinitis and asthma [7,14], as well as other chronic conditions [15]. To measure perceived control, we used the 8-item PCRQ. PCRQ scores range from 8 to 40, with higher scores reflecting greater perceived control of rhinitis. We previously validated this instrument in a cross-sectional analysis of our baseline data [7]. Similar instruments, upon which it is based, have also been validated for use in asthma [14].
Disability, health care utilization, and medication costs
Other rhinitis-specific health outcomes assessed included job effectiveness and restricted activity over the 4 weeks prior to interview, and physician visits and medication costs over the 12 months prior to interview. Job effectiveness, in terms of the specific impact of the subject's nasal condition, was assessed on a scale of 0 to 100% (0% meaning unable to work at all and 100% meaning work not affected). Restricted activity was reported as the number of days restricted due to subject's medical condition. Physician visits were reported as the number of visits made specifically for a nasal condition, excluding routine visits for allergy desensitization. Out-of-pocket health costs were assessed separately for prescription and over-the-counter rhinitis medications. Subjects were asked to choose from four possible cost ranges: 'less than $10', '$10 to $99', '$100 to $1,000', or 'more than $1,000'. For the purpose of our analysis, we assigned subjects to the median value for each response category (e.g. $45 for '$10 to $99'), using $1000 for the highest category. A single cost variable was then calculated by summing costs for prescription and over-the-counter medications.
Statistical analyses
Comparisons between subjects who completed a follow-up interview (n = 69) with those who had not (n = 40) were made using t test for continuous variables, Chi-square test for dichotomous variables, and Chi-square test for trend for ordinal variables. Mean (±SD) summary scores were calculated for each health status measure at baseline and 24-month follow-up. Change was calculated by subtracting baseline scores from scores at follow-up and expressed as standardized response means (SRM). The SRM equates to the overall observed change divided by the standard deviation of the individual observed differences in score for the entire group. For the RSDI, a negative change in score reflects improved HRQL over time, whereas a positive change in score reflects worsened HRQL. To evaluate for changes in the RSDI within individuals, we used the paired t test.
We examined responsiveness of the RSDI in two separate ways. First, we evaluated change in the RSDI score relative to change in the other health status measures delineated above (RSS-4, SF-12 PCS, CES-D, and PCRQ), treating all variables as continuous. Pearson correlations were used to make these comparisons, as changes in scores approximated normal distributions for all health measures studied. As an alternative analysis, we categorized subjects into three groups, 'better', 'same', or 'worse', based on change in each health status measure. We used a change quantified as one standard error of measurement (SEM) or greater as the criterion for a difference great enough to be consistent with a minimal clinically important difference (MCID). Evidence by Wywrich and others supports use of one SEM (calculated as the standard deviation of the instrument multiplied by the square root of one minus its reliability coefficient) as an approximation of the MCID [16]. By inference, this should also be a reasonable gauge of a substantive difference in score over time. Thus using this categorization scheme, we calculated SEM-based MCIDs for the RSS-4, SF-12 PCS, CES-D, and PCRQ. Based on these cut-offs, we defined subjects as 'same' (absolute change in score <1 SEM), 'better' (score improved by ≥1 SEM), or 'worse' (score worsened by ≥1 SEM) for each measure. Mean change in RSDI was compared among the groups, first using one-way ANOVA to detect an overall difference, then using Tukey's t test to perform multiple comparisons (when ANOVA was significant at p < 0.05). We repeated this analysis for each health status measure to which the RSDI was compared (RSS-4, SF-12 PCS, CES-D, and PCRQ).
Finally, we used linear regression to evaluate the relationship between change in HRQL, as measured by the RSDI, and change in rhinitis-specific health outcomes (disability, health care utilization, and medication costs). We first evaluated each health outcome in a bivariate model with change in RSDI score as the independent variable. We then performed multiple linear regression adding baseline RSS-4, SF-12 PCS, CES-D, and PCRQ as covariates in the model, in order to take into account each subject's health status at baseline. Effect estimates are expressed as change in the health outcome variable (days of restricted activity, percent change in job effectiveness, number of visits, or dollars spent for medications) per 1.0 SRM change in the RSDI score. All analyses were performed using SAS System for Windows release 8.02.
Results
Subjects characteristics
Overall, the 69 subjects studied were predominately female, white (non-Hispanic), and well-educated, with predominately moderate to severe self-rated rhinitis (Table 1). Baseline characteristics of those subjects re-interviewed were not significantly different that those who were not, with the exception of income. Subjects not re-interviewed were more likely to have an annual family income of less than $40,000.
Table 1 Subject characteristics of 109 adults with rhinitis interviewed at baseline
Re-interviewed at 24 months
Baseline characteristics Yes (n = 69) No (n = 40) p value
Age, mean ± SD 40 ± 8.2 37 ± 8.6 0.16
Female, n (%) 44 (64) 30 (75) 0.23
White (non-Hispanic), n (%) 50 (72) 30 (75) 0.77
Education, n (%) 0.36
High school or less 10 (14) 8 (20)
Some college 47 (68) 27 (68)
Graduate degree 12 (17) 5 (13)
Married, n (%) 44 (64) 23 (58) 0.52
Income <$40,000/y, n (%) 10 (14) 12 (30) 0.05
Smoking status, n (%) 0.46
Never 43 (62) 22 (55)
Former 15 (22) 10 (25)
Current 11 (16) 8 (20)
Self-rated severity of rhinitis, n (%) 0.95
Mild 24 (35) 14 (34)
Moderate 35 (51) 19 (48)
Severe 10 (14) 7 (18)
Change in HRQL and health status
The change in HRQL over follow-up approximated a normal distribution, with the majority of subjects demonstrating a change of less than 0.5 SRM (Figure 1). The RSDI decreased by 0.5 SRM or more (better HRQL) in 13 (19%) subjects and increased by 0.5 SRM or more (worse HRQL) in 17 (25%). Overall, RSDI scores for the group as a whole did not change significantly (Table 2). Similarly, no statistically significant differences were observed in any of the RSDI subscales.
Figure 1 Change in health-related quality of life (HRQL) over 24 months in 69 adults with rhinitis. Width of each bar represents one standardized response mean (SRM) calculated as the mean observed change divided by the standard deviation of the observed change. A negative change in RSDI score reflects better HRQL, whereas a positive change in score reflects worse HRQL.
Table 2 Change in health-related quality of life and health status among 69 adults with rhinitis
Measure Baseline (mean ± SD) 24 month follow-up (mean ± SD) Observed change* (mean ± SE) Standardized response mean† p value‡
RSDI – Total 23.3 ± 23.4 24.6 ± 24.0 1.3 ± 2.7 0.06 0.63
Physical 11.2 ± 10.6 10.7 ± 9.6 -0.4 ± 1.2 -0.04 0.72
Functional 5.6 ± 7.0 6.5 ± 7.5 0.9 ± 0.9 0.13 0.28
Emotional 6.6 ± 7.7 7.3 ± 8.4 0.8 ± 1.0 0.10 0.42
RSS-4 6.9 ± 4.6 5.4 ± 3.9 -1.5 ± 0.5 -0.31 0.01
CES-D 10.5 ± 10 11.5 ± 9.9 1.0 ± 1.3 0.09 0.43
SF-12 PCS 48.2 ± 8.6 47.8 ± 9.0 -0.5 ± 1.0 -0.05 0.65
PCRQ 27.7 ± 4.8 28.8 ± 5.2 1.2 ± 0.7 0.21 0.09
RSDI = Rhinosinusitis Disability Index; RSS-4 = 4-item Rhinitis Symptom Score; CES-D = Center for Epidemiologic Studies Depression scale; SF-12 PCS = Short Form 12 Physical Component Summary; PCRQ = Perceived Control of Rhinitis Questionnaire. * Observed change = (score at follow-up) - (score at baseline). † Standardized response mean = [(score at follow-up) - (score at baseline)/(SD of observed change)]. ‡ Paired t test
Changes in symptom severity, physical functioning, mood, and perceived control also approximated normal distributions. Rhinitis symptom severity, as measured by the RSS-4, improved by a small, though significant, increment (Table 2). The remaining health status measures (SF-12 PCS, CES-D, and PCRQ) demonstrated no statistically significant changes at the group level.
Responsiveness to changes in health status
Change in the RSDI at the individual level correlated moderately well with changes in the RSS-4, SF-12 PCS, CES-D, and PCRQ. Moreover, changes were in the anticipated directions (Table 3). Specifically, better HRQL, as measured by the RSDI, was associated with lower symptom severity, greater physical functioning, better mood, and greater perceived control. Similar correlations were observed for each of the RSDI subscales.
Table 3 Correlations between change in the RSDI and change in other health status measures
Δ RSS-4 Δ SF-12 PCS Δ CES-D Δ PCRQ
Δ RSDI – Total 0.38 (0.16, 0.57) -0.39 (-0.57, -0.16) 0.37 (0.15, 0.56) -0.37 (-0.56, -0.14)
Physical 0.32 (0.09, 0.52) -0.37 (-0.55, -0.14) 0.31 (0.08, 0.51) -0.30 (-0.50, -0.07)
Functional 0.39 (0.17, 0.58) -0.45 (-0.62, -0.23) 0.39 (0.17, 0.57) -0.39 (-0.57, -0.17)
Emotional 0.33 (0.10, 0.52) -0.23 (-0.44, 0.01) 0.31 (0.08, 0.51) -0.32 (-0.51, -0.09)
Values represent Pearson correlation coefficients (95% confidence interval). RSDI = Rhinosinusitis Disability Index; RSS-4 = 4-item Rhinitis Symptom Score; CES-D = Center for Epidemiologic Studies Depression scale; SF-12 PCS = Short Form 12 Physical Component Summary; PCRQ = Perceived Control of Rhinitis Questionnaire. Δ = Change over 24 month follow-up.
We reanalyzed these relationships, categorizing subjects based on change in health status as measured by the RSS-4, SF-12 PCS, CES-D, and PCRQ. Mean RSDI scores improved (negative change in score) in subjects categorized as 'better' with respect to symptom severity (RSS-4), physical functioning (SF-12 PCS), mood (CES-D), and perceived control (PCRQ) (Figure 2). Similarly, mean RSDI scores worsened (positive change in score) in subjects categorized as 'worse' according to the other health status measures. In all cases, observed changes in the RSDI were statistically significant compared to subjects categorized as the 'same' relative to baseline.
Figure 2 Responsiveness of RSDI to change in other health status measures. Bars represent mean change in RSDI score categorized by change in health status (better/same/worse, see legend above). A negative bars (decrease in RDSI) reflect better health-related quality of life (HRQL), whereas a positive bars (increase in RSDI) reflect worse HRQL. Symptom severity, physical functioning, psychological mood, and perceived control of rhinitis were assessed using the RSS-4, SF-12 PCS, CES-D, and PCRQ, respectively (see Methods).
These same results are presented again in an alternative format in Figure 3, showing categorical change in the RSDI versus categorical change in the other health status measures. Highly discordant change between measures (for example, 'better' versus 'worse') was observed in approximately one in every 10 subjects: 4 (6%) subjects for the symptom severity (RSS-4), 6 (9%) subjects for physical functioning (SF-12 PCS), 9 (13%) subjects for mood (CES-D), and 7 (10%) subjects for perceived control (PCRQ).
Figure 3 Frequency counts for categorical change between the RSDI and other health status measures. Values represent the number of subjects within each category. Shaded cells indicate highly discordant change between measures. Overall, such discordance occurred in approximately one in every 10 subjects. Change in the RSDI was highly discordant in 6% of subjects for the RSS-4, 9% of subjects of the SF-12 PCS, 13% of subjects for the CES-D, and 10% of subjects for the PCRQ. For definitions of abbreviations see Table 2.
Relationship with self-reported health outcomes
Change in HRQL, as measured by the RSDI, was significantly associated with change in days of restricted activity, altered job effectiveness, and incremental medication costs, but was not associated with differing frequency of physician visits relative to baseline (Table 4). After adjusting for baseline health status (RSS-4, SF-12 PCS, CES-D, and PCRQ), these observed associations for activities, effectiveness and medication costs remained statistically significant. In these multivariate analyses, an increase in RSDI score of 1.0 SRM (22.3 points) was associated with an increase of nearly 2.5 days of restricted activity per month, a decrease of >4% in job effectiveness, and an increase of >$78 spent on rhinitis medications per year.
Table 4 Relationship between change in RSDI and change in rhinitis-specific health outcomes
Change in RSDI as a predictor of change in health outcome
Unadjusted model Adjusted model
Dependent variable Mean change ± SD β ± SE p value β ± SE p value
Restricted activity (days per month) 2.4 ± 7.3 2.48 ± 0.84 <0.01 2.48 ± 0.98 0.01
Job effectiveness* (% effectiveness) -2.4 ± 12.8 -4.78 ± 2.03 0.02 -4.27 ± 1.95 0.03
Physician visits (# of visits per yr) -0.1 ± 3.1 0.41 ± 0.46 0.38 0.36 ± 0.47 0.45
Cost of medications ($ per year) 4.9 ± 309 87.44 ± 36.16 0.02 78.52 ± 38.73 0.05
Effect estimates are reported per 1.0 SRM change in the RSDI score. Adjusted model includes baseline symptom severity (RSS-4), physical functioning (SF-12 PCS), psychological distress (CES-D), and perceived control of disease (PCRQ). * Data not available for 24 subjects (9 unemployed, 11 house keeping, 2 attending school, 1 retired, 1 other).
Discussion
In this study, we demonstrate that the RSDI is responsive to changes in health status over time, and thus can be used to measure longitudinal change in HRQL in rhinitis. Change in HRQL, as measured by the RSDI, correlated in the expected ways with changes in symptom severity, physical functioning, mood, and perceived control of symptoms. In addition, we found that change in HRQL was associated with changes in rhinitis-specific health outcomes, specifically days of restricted activity, job effectiveness, and medication costs, even after controlling for health status at baseline.
Although several measures of HRQL have been developed for rhinitis and sinusitis, few have undergone rigorous psychometric testing [2]. In order for an instrument to be useful, it should be valid, reliable, and responsive. A systematic review by Linder et al. identified 16 instruments used to measure HRQL in sinusitis, of which only three instruments met basic requirements for validity, reliability, and responsiveness: the Chronic Sinusitis Survey – Duration-based (CSS-D), the Rhinosinusitis Outcome Measure-31 (ROM-31), and the SNOT-16 [17]. Additionally, the Juniper RQLQ has also been shown to demonstrate strong measurement properties, particularly in patients with allergic rhinitis [18].
This study focuses on demonstrating the responsiveness of the RSDI, which has yet to be reported. We believe that this is important because the RSDI is useful across a range of rhinologic conditions, is reasonably short in length, and is structured in a way that facilitates telephone administration. These attributes that make the RSDI particularly suitable for repeated administration in longitudinal survey research where the ability to measure change is often desired.
Responsiveness can be determined in a number of ways, but must include some type of longitudinal assessment [19]. There are two general approaches to interpreting change in HRQL, a 'distribution-based' approach (also referred to as 'internal responsiveness') and an 'anchor-based' approach (also referred to as 'external responsiveness') [20,21].
The 'distribution-based' approach relies entirely on the statistical distribution of the results, using effect size or SRM as a method for assessing change. For our initial analysis, we adopted this approach to summarize global changes in the RSDI, expressing the difference in scores as standardized response means (Figure 1), and testing the difference using paired t test (Table 2). The main criticism of relying solely on a 'distribution-based' approach is that there is no standard by which to judge whether the changes observed are in fact important to the patient or clinically meaningful. Often this approach is used in clinical trials involving an intervention that is believed to be efficacious. The RSDI has only been used in two clinical trials, neither of which purported to directly assess the responsiveness of the instrument. In one trial evaluating the efficacy of hypertonic saline nasal irrigation, RSDI scores improved by 6.0 to 15.5 points in the treated group compared with controls [22]. In our analysis, we used 0.5 SRM as a measure of change, which corresponds to approximately 11 points on the RSDI. Another trial, which evaluated the utility of a treatment protocol for rhinosinusitis, showed no statistical difference in RSDI scores between those who felt improved and those who did not, although there were measurable differences in other clinical outcomes [23].
Unlike the 'distribution-based' approach, 'anchor-based' approaches attempt to compare, or anchor, changes in a measure to some external criterion. This can be particularly troublesome for HRQL instruments, where no gold standard exists. In this case, the instrument must be compared to other established measures that assess related constructs. Cross-sectional studies in rhinitis have demonstrated that objective clinical measures, such as CT scoring and nasal endoscopy, correlate poorly with symptom severity and quality of life [24,25]. Therefore, it is difficult to argue that responsiveness should be measured in these terms. Instead, we chose to measure the responsiveness of the RSDI relative to other related health status measures to which it has already been shown to correlate with cross-sectionally.
In a previous analysis, we demonstrated that HRQL in rhinitis, as measured by the RSDI, correlates with measures of symptom severity, physical functioning, mood, and perceived control assessed simultaneously [7]. In this current analysis, we studied a subset of these subjects for whom longitudinal data was available. Because subjects were studied 2 years after baseline, enough time had elapsed for health status to have improved in some (either spontaneously or due to medical intervention) and to have worsened in others (due the disease progression or under-treatment). We measured responsiveness of the RSDI to these changes in health status in two ways. We did this first, by correlating change in the RSDI with change in other health status measures on a continuous scale (Table 3) and second, by dividing subjects into categories (better, same, worse) based on significant changes in health status and comparing differences in RSDI scores between categories (Figure 2). Although these analyses may be statistically similar, they are conceptually different. Using both of these methods, we found that the RSDI performed in the hypothesized manner. Additionally, we also compared categorical change in the RSDI versus categorical change in the other health status measures (Figure 3). Presented in this way, we found extreme discordance between measures in only a minority of cases, also consistent with an overall responsiveness to change.
Finally, to provide further support for utility of the RSDI as a measure of change, we used the RSDI to evaluate the relationship between change in HRQL and change in specific health outcomes that we believed to be most relevant to patients suffering from rhinitis. We found that, even after controlling for baseline health status, change in the RSDI score was indeed an independent predictor of change in days of restricted activity, job effectiveness, and medication costs.
We failed to demonstrate, however, that the RSDI is responsive to change in number of physicians visits. This negative finding could be explained by a number of reasons. First, there was little change in the number of visits observed for the group as a whole. Second, physician visits were queried over the 12 months prior to interview as opposed to the RSDI, which was has a recall period of only 4 weeks. Finally, patients with rhinitis often self-manage their own symptoms, and therefore fluctuations in disease severity may not be necessarily be captured by physician visits.
Due to the lack of consensus on the single, best method for determining responsiveness [19], we used a combination of approaches to try to assess whether the RSDI is sensitive to meaningful changes in other measures that should correlate with HRQL. We recognize that the strength of our conclusions can only be as strong as the health status measures we have chosen to use for comparison. This limitation, however, is inherent to any study which proposes to measure the responsiveness of HRQL instruments, given that no gold standard exists.
Some might contend that because health status did not change for the study population as a whole over the 2 years, then what we are in fact measuring is longitudinal construct validity, rather than true responsiveness. This particular distinction remains a topic of ongoing debate [19]. Certain authors have argued that responsiveness represents a psychometric property separate from validity [26,27], whereas others believe that responsiveness should be treated as a form of longitudinal validity [28,29]. When dealing with the measurement of HRQL, this distinction becomes even less clear. Bearing these limitations in mind, the longitudinal characteristics of the RSDI demonstrated in this study meet and exceed most performance expectations for an evaluative HRQL instrument.
Conclusion
In summary, we conclude that RSDI is responsive to changes in HRQL as indicated by its correlation with other health status measures and rhinitis-specific outcomes measured longitudinally. Because the treatment for rhinosinusitis is based primarily on symptoms and their impact on the individual, it is important to have quantifiable measures that are sensitive to change in health status. Based on the responsiveness of the RSDI we observed, combined with its ease of administration and applicability, this instrument should be considered for future use in other clinical studies of rhinitis and sinusitis.
List of abbreviations
CCS-D: Chronic Sinusitis Survey – Duration-based
CES-D: Center for Epidemiologic Studies depression scale
HRQL: Health-related quality of life
MCID: Minimal clinically important difference
PCRQ: Perceived Control of Rhinitis Questionnaire
ROM-31: 31-item Rhinosinusitis Outcome Measure
RQLQ: Rhinoconjunctivitis Quality of Life Questionnaire
RSDI: Rhinosinusitis Disability Index
RSS-4: 4-item Rhinitis Symptom Score
SEM: Standard error of measurement
SF-12 PCS: Short Form 12 physical component summary
SNOT-16: 16-item Sino-Nasal Outcome Test
SRM: Standardized response mean
Authors' contributions
HC conceived and designed the study, performed the statistical analysis, and drafted the manuscript. PB contributed to conception and design of the study, provided the data on which this analysis was based, participated in the interpretation of data, and made substantial revisions to the manuscript. PK provided expertise on psychometric analysis, participated in the interpretation of data, and critically reviewed the final manuscript. SS provided statistical consultation and critically reviewed the final manuscript.
Supplementary Material
Additional file 1
The Rhinosinusitis Disability Index (RSDI) Domains and Items.
Click here for file
Acknowledgements
Funded by National Institutes of Health R01 ES10906. Dr. Chen also funded by F32 HL077994
==== Refs
Kremer B Quality of life scales in allergic rhinitis Curr Opin Allergy Clin Immunol 2004 4 171 176 15126937 10.1097/00130832-200406000-00006
Linder JA Atlas SJ Health-related quality of life in patients with sinusitis Curr Allergy Asthma Rep 2004 4 490 495 15462717
Juniper EF Guyatt GH Development and testing of a new measure of health status for clinical trials in rhinoconjunctivitis Clin Exp Allergy 1991 21 77 83 2021881
Anderson ER Murphy MP Weymuller EAJ Clinimetric evaluation of the Sinonasal Outcome Test-16. Student Research Award 1998 Otolaryngol Head Neck Surg 1999 121 702 707 10580223 10.1053/hn.1999.v121.a100114
Benninger MS Senior BA The development of the Rhinosinusitis Disability Index Arch Otolaryngol Head Neck Surg 1997 123 1175 1179 9366696
Senior BA Glaze C Benninger MS Use of the Rhinosinusitis Disability Index (RSDI) in rhinologic disease Am J Rhinol 2001 15 15 20 11258649
Chen H Katz PP Eisner MD Yelin EH Blanc PD Health-related quality of life in adult rhinitis: the role of perceived control of disease J Allergy Clin Immunol 2004 114 845 850 15480325 10.1016/j.jaci.2004.07.008
Blanc PD Trupin L Eisner M Earnest G Katz PP Israel L Yelin EH The work impact of asthma and rhinitis: findings from a population-based survey J Clin Epidemiol 2001 54 610 618 11377122 10.1016/S0895-4356(00)00349-8
Wasserfallen JB Gold K Schulman KA Baraniuk JN Development and validation of a rhinoconjunctivitis and asthma symptom score for use as an outcome measure in clinical trials J Allergy Clin Immunol 1997 100 16 22 9257782 10.1016/S0091-6749(97)70189-1
Ware JJ Kosinski M Keller SD A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity Med Care 1996 34 220 233 8628042 10.1097/00005650-199603000-00003
Radloff LS The CES-D Scale: a self-report depression scale for research in the general population Appl Psychol Meas 1977 1 385 401
Breslau N Depressive symptoms, major depression, and generalized anxiety: a comparison of self-reports on CES-D and results from diagnostic interviews Psychiatry Res 1985 15 219 229 3862157 10.1016/0165-1781(85)90079-4
Orme JG Reis J Herz EJ Factorial and discriminant validity of the Center for Epidemiological Studies Depression (CES-D) scale J Clin Psychol 1986 42 28 33 3950011
Katz PP Yelin EH Eisner MD Blanc PD Perceived control of asthma and quality of life among adults with asthma Ann Allergy Asthma Immunol 2002 89 251 258 12269644
Nicassio PM Wallston KA Callahan LF Herbert M Pincus T The measurement of helplessness in rheumatoid arthritis. The development of the arthritis helplessness index J Rheumatol 1985 12 462 467 4045844
Wyrwich KW Tierney WM Wolinsky FD Further evidence supporting an SEM-based criterion for identifying meaningful intra-individual changes in health-related quality of life J Clin Epidemiol 1999 52 861 873 10529027 10.1016/S0895-4356(99)00071-2
Linder JA Singer DE Ancker M Atlas SJ Measures of health-related quality of life for adults with acute sinusitis. A systematic review J Gen Intern Med 2003 18 390 401 12795739 10.1046/j.1525-1497.2003.20744.x
Juniper EF Measuring health-related quality of life in rhinitis J Allergy Clin Immunol 1997 99 S742 9 9042066
Terwee CB Dekker FW Wiersinga WM Prummel MF Bossuyt PM On assessing responsiveness of health-related quality of life instruments: guidelines for instrument evaluation Qual Life Res 2003 12 349 362 12797708 10.1023/A:1023499322593
Lydick E Epstein RS Interpretation of quality of life changes Qual Life Res 1993 2 221 226 8401458 10.1007/BF00435226
Husted JA Cook RJ Farewell VT Gladman DD Methods for assessing responsiveness: a critical review and recommendations J Clin Epidemiol 2000 53 459 468 10812317 10.1016/S0895-4356(99)00206-1
Rabago D Zgierska A Mundt M Barrett B Bobula J Maberry R Efficacy of daily hypertonic saline nasal irrigation among patients with sinusitis: a randomized controlled trial J Fam Pract 2002 51 1049 1055 12540331
Kinney WC Rhinosinusitis treatment protocol: changing provider habits in primary care Otolaryngol Head Neck Surg 2002 126 614 622 12087327 10.1067/mhn.2002.125301
Birch DS Saleh HA Wodehouse T Simpson IN Mackay IS Assessing the quality of life for patients with chronic rhinosinusitis using the "Rhinosinusitis Disability Index" Rhinology 2001 39 191 196 11826687
Krouse JH Computed tomography stage, allergy testing, and quality of life in patients with sinusitis Otolaryngol Head Neck Surg 2000 123 389 392 11020173 10.1067/mhn.2000.109476
Guyatt GH Deyo RA Charlson M Levine MN Mitchell A Responsiveness and validity in health status measurement: a clarification J Clin Epidemiol 1989 42 403 408 2659745 10.1016/0895-4356(89)90128-5
Puhan MA Bryant D Guyatt GH Heels-Ansdell D Schunemann HJ Internal consistency reliability is a poor predictor of responsiveness Health Qual Life Outcomes 2005 3 33 15877824 10.1186/1477-7525-3-33
Hays RD Hadorn D Responsiveness to change: an aspect of validity, not a separate dimension Qual Life Res 1992 1 73 75 1301117 10.1007/BF00435438
Lindeboom R Sprangers MA Zwinderman AH Responsiveness: a reinvention of the wheel? Health Qual Life Outcomes 2005 3 8 15691385 10.1186/1477-7525-3-8
|
16277662
|
PMC1308842
|
CC BY
|
2021-01-04 16:38:13
|
no
|
Health Qual Life Outcomes. 2005 Nov 8; 3:68
|
utf-8
|
Health Qual Life Outcomes
| 2,005 |
10.1186/1477-7525-3-68
|
oa_comm
|
==== Front
Health Qual Life OutcomesHealth and Quality of Life Outcomes1477-7525BioMed Central London 1477-7525-3-711628588410.1186/1477-7525-3-71ResearchThe visual analog rating scale of health-related quality of life: an examination of end-digit preferences Shmueli Amir [email protected] Department of Health Management, The Hebrew University, POB 12272 Jerusalem, Israel2005 14 11 2005 3 71 71 19 9 2005 14 11 2005 Copyright © 2005 Shmueli; licensee BioMed Central Ltd.2005Shmueli; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The Visual Analog Scale (VAS) has been extensively used in the valuation of health-related quality of life (HRQL). The objective of this paper is to examine the measurement error (rounding) explanation for the higher prevalence of VAS scores ending with a zero, and to provide an alternative interpretation.
Methods
The analysis is based on more than 4,500 reported VAS valuations of own HRQL, included in two Israeli health surveys (1993 and 2000). Bivariate and logistic regression analyses are used.
Results
The results show that reporting VAS scores ending with a 0 (...-20, ..0,10,20.....) decreases and scores ending with a 5 (...-15,-5,5,15,25,...) and with any other integer (...-12, -11,...1,2,...,92,..99) increases as VAS scores depart from 50, particularly when increasing up to 100. This pattern remains after controlling for personal characteristics determining the level of VAS.
Discussion
Rounding true HRQL to the nearest 10's or 5's cannot explain the specific pattern found. It is suggested that this pattern corresponds to a S-shaped value function, where individuals tend to evaluate their HRQL as "gains" or "losses" relative to a reference point evaluated at 50. This particular reference score originates from being a traditional "passing threshold" and the scale's midpoint. Several implications of this interpretation to the measurement of HRQL are discussed.
Visual Analog ScaleEnd-digit preferenceHealth-Related Quality of Life
==== Body
Background
Because of its simplicity and practical applicability, the Visual Analog Scale (VAS) has been widely used to elicit individuals' health value functions, either through measuring preferences for specific health states [1,2] or through evaluating their own health-related quality of life (HRQL) [3-5]. Recently, several studies examined the theoretical foundation of the VAS in relation to Von-Neumann-Morgenstern utility theory, and explored certain measurement problems such as end of scale aversion and spacing-out bias [2,6,7].
The present study focuses on end-digit preferences of the VAS scores, used to evaluate own HRQL. End-digit preference in reporting is not new, it was detected in 1940 in reporting age, and was later detected in blood pressure measurement, birth weight recording, and estimated gestational age [8,9]. The relative concentration of reported VAS scores ending with a 0 has been interpreted, as was done in the just mentioned studies in other contexts, as measurement errors, where people "round" their valuations to the nearest 10, while true HRQL is a continuous variable. However, it is shown that the closer the score is to 100 (perfect health), the higher the relative frequency of scores ending with an integer other than 0. Consequently, a different interpretation of the results is based on the assumption that no rounding is used, and respondents deliberately choose scores ending with 0, 5 or other integer to accurately reflect their HRQL. That interpretation, which implies an underlying S-shaped relationship between the VAS and true HRQL, is discussed.
Methods
The survey data
The data used in this study comes from two full sit-down health surveys – conducted in 1993 and in 2000 – of the Israeli Jewish urban population aged 45–75. Stratified (by settlement size) samples were used to represent the population studied. The 1993 survey included 1,999 individuals, while the 2000 survey included 2,505 individuals (for more details see [10]). Preliminary analysis showed that similar results (see below) are obtained for both years. Consequently, the final analysis reported below included the pooled two-year sample.
The measurement of HRQL by the VAS
In both surveys, HRQL was valued in the following way: A card with a vertical scale ranging from -100 to +100, with unit marks (1s) and numbers appearing every five scores (at 5s and 10s), was presented to the respondents. The respondents were told that zero signifies HRQL associated with death, and 100 – HRQL associated with perfect health (regardless of age). The interviewers added that negative values are possible, meaning HRQL worse than that associated with death. The respondents were asked to report verbally the number on the above scale, which represents their general HRQL during the previous month.
The statistical analysis
Bivariate and multivariate logistic regression analyses were used to show that the probability of VAS scores ending with an integer other than 0 or 5 differs in different ranges of scores. One may argue that such a pattern originates from the different characteristics of the respondents who chose different score ranges rather than from the scale itself. For example, persons enjoying very high HRQL might tend to report scores not ending with 0 or 5 more than other individuals. To examine that argument, selected personal characteristics, which are likely to affect the reported VAS score, were controlled for. These characteristics included: economic status (a set of 4 dummy variables representing the five categories: excellent, very good, good, fair and poor), ethnic origin (a set of 3 dummy variables representing the four categories: Asia-Africa, Europe-America, Israel and post 1990 immigrants from the former USSR), years of education, gender and age.
Results
Figure 1 presents the distribution of VAS scores for the two years combined. The minimum score reported was -20. The mode of the distribution is 71–80, and the distribution is skewed to the left. For later reference, note the somewhat outstanding high frequency of the category 41–50.
Figure 1 The distribution of VAS scores.
Overall, 89% reported scores ending with 0 ("10s", -20, -10, 0, 10, ....,100), 9% reported scores ending with 5 ("5s", -15, -5, 5, 15,....95), and 2% reported a score ending with another integer ("1s" or all other scores, namely, -19, -18, ....1, 2, .....49, 51,....83, ...99). Figure 2 presents the (stacked) percentages of scores ending with 0, 5 and another integer by valuation categories, for the two years combined. The Figure shows that in the categories "< = 0" and "41–50", 98–100% of the scores are multiples of 10. In other words, all the scores below or equal to zero, are -20, -10 or 0. Similarly, 98% of the scores between (including) 41 and 50, equal, actually, 50.
Figure 2 Proportions of scores ending with 0, 5, and other digits by valuation score.
Once the score is greater than 0 and lower than 40, or greater than 50, the percentages of 10s drop, and the proportions of scores ending with a 5 or another integer increase. For example, in the category "1–10", 85% of the scores equal 10, and 15% equal 5. This trend is more pronounced for scores greater than 50: in the 51–60 category, 94% of the scores equal 60, 4.5% equal 55, and 1.2% equal one of the remaining scores. In the upper category (91–100), 76% are equal to 100, 15% chose 95, and more than 9% are other scores in the category (92, 93, etc). The almost-steady increase in the proportions of scores not ending with a 0 or with a 5 is clear for scores higher than 40.
In order to test statistically the hypothesis that the proportion of scores ending with 0 is constant across the score-level categories (as is expected if just rounding was the issue), a logistic regression of the probability of a score ending with a 0 was run on the 11 score categories' dummy indicators. The Likelihood Ratio statistic, testing the hypothesis that all the score-level categories effects are equal, was 276.7 (DF = 10), which indicates that the hypothesis is rejected. Namely, the probabilities of a score ending with a 0 differ across the score-level categories. Controlling for the personal characteristics did not change the results. This means that the variable proportions of scores ending with a 0 and 5 (and hence of all other scores) by score range does originate from the VAS properties and not from the respondents' differing characteristics determining their score category.
Figure 3 shows the same results in terms of deviations of the actual number of cases from the expected ones, under a uniform distribution with rates equal the total's proportions of scores ending with a 0, 5 and other scores, by valuation score. The actual number of 10s is greater than the expected one in all scores up to score 80. It increases up to score range 41–50, and then drops. For score ranges 81–90 and 91–100, the actual number of 10s is smaller than the expected one. The deviations in the numbers of scores ending with a 5 by score range are almost an exact mirror image of the deviations of the number of 10s. They are negative and decreasing up to score of 50, and then negative and increasing up to 70, they continue to increase up to 90, and drop in the score range 91–100. The pattern of the deviations in the number of scores not ending with a 0 or 5 is similar to the one of the deviations in the number of scores ending with a 5, but smoother. Also, the deviations increase steadily from 71–80 up to 91–100.
Figure 3 Actual minus expected (under a uniform distribution) number of cases ending with 0, 5, or other digit by valuation score.
Discussion
Health-related quality of life (HRQL) is a latent continuous construct. VAS scores provide a measure of that unobservable variable. Much experience has shown that the VAS is easy to obtain, and respondents have no problems in scoring. The patterns presented above imply a particular relationship between the VAS reports and HRQL. The results, as shown in Figure 2 in particular, indicate a distinctive role for the score of 50. First, it is the score ending with a 0 with the largest concentration of responses. Second, disregarding for a moment negative scores, 98% of the individual scores within its neighboring score range are concentrated at its value, the highest concentration across all score ranges. Consequently, the percentage of scores ending with an integer other than 0, increases as the score range furthers away from 50, upward and downward. The score of 50 may be thus considered as an empirical reference or benchmark score (see below for an interpretation).
Figure 4 presents the relationship between the VAS scores and true HRQL, which emerges from the above results. In Figure 4, qp, the true HRQL for which the VAS score is 50, is the true reference point of the scale. Levels of HRQL in the neighborhood of qp are not that different, leading respondents with HRQL in this range to round their VAS scores to 50. At that point, the curve is relatively vertical, since true HRQL is relatively constant. Scores in this range ending with integers other than 0 (say 41, 42,...., 49, 51,...59) indicate approximately the same level of HRQL – qp, so they are almost not reported. It takes 10 units of the scale (say 40, 60) to indicate a different level of HRQL.
Figure 4 The implied shape of VAS scores as a function of true health-related quality of life.
The higher the VAS score (from 50 and up), the larger the difference in true HRQL (measured horizontally in figure 4) for a given difference in VAS (measured vertically). For VAS>91, each additional point on the score signifies relatively dramatically higher true HRQL. In this category every point is significant since HRQL is rapidly changing. For that reason, scores ending with an integer other than 0 are most frequent in this score range. Graphically, this translates into the curve being relatively flat (put inversely, the VAS is relatively constant over a relatively wide range of HRQL), and the curve is concave from below for HRQL values higher than qp.
A similar relationship between the VAS and HRQL holds for 0<HRQL< qp, with the curve being flatter for HRQL approaching that of death, so that the curve is convex from below for HRQL lower than qp.
A second threshold in the relationship is at HRQL = death, for which VAS = 0. As was argued above, the VAS for true HRQL worse than death is quite insensitive to the precise level of HRQL (scores ending with 0, the curve being graphically steep), and it takes differences of 10 points to indicate different levels of true HRQL. This threshold is defined, however, by the instructions. While the range of HRQL worse than death is extremely interesting and important, the analysis of this range is not very reliable, as only 10 persons (out of 4,504) reported negative scores on the VAS.
The value function in Figure 4 might represent valuation in a way similar to the one on which prospect theory is based [11]. Abstracting from uncertainty issues, prospect theory suggests that individuals do not evaluate states (e.g. levels of wealth) in their absolute value (as in Friedman-Savage utility theory) but as deviations (monetary gains or losses) relative to some reference point (e.g. the present level of wealth). Furthermore, the value function is concave (diminishing marginal value) for positive deviations (gains) and convex (increasing marginal value) for negative ones. Finally, the value function is steeper at each level of negative deviation than at the positive equal deviation. The value function depicted in Figure 4 following the empirical characteristics of the VAS reports, matches these characteristics. For non-negative HRQL, individuals evaluate their HRQL in relation to the reference value qp, which is the level of HRQL evaluated as 50. For HRQL better than qp, individuals consider the difference (HRQL-qp) as a "gain", and report a VAS value accordingly, with diminishing marginal value. For HRQL worse than qp, individuals consider the difference (HRQL-qp) as a "loss", and report a VAS value accordingly, with increasing marginal value. As is clear from Figure 2, the value function in Figure 4 is steeper for negative deviations (HRQL< qp) than for equal but positive deviations (HRQL> qp).
The distinctive role of the reference point qp evaluated by 50 is suggested by the data. Nevertheless, what can be the interpretation of these values? Two explanations can be offered. First, in Israel, as in many other education systems, the evaluation of the pupils' achievements is done by a grade on a 0–100 scale. On this scale, a grade of 50 is usually considered a "passing grade", where lower grades indicate a failure (in Israel, failing grades are commonly called "negative grades", reflecting the "loss" with respect to the passing grade 50 as a reference point recorded as 0). A second explanation sees 50 as simply the midpoint on the positive 0–100 scale. The psychometric importance of scales' midpoint is well known, e.g., the "midpoint bias", where (too) many respondents tend to choose the mid category from among an odd number of options.
The significance of qp evaluated as the mid-scale 50 is closely related to the "bisection procedure", where respondents matched, by a sequence of bisections, a number (magnitude) to brightness and loudness. This procedure was found to agree fairly closely with matching done by magnitude estimation (where numbers are directly matched to stimuli). Furthermore, for the magnitude estimation procedure, it is clearly stated that: " [....] stimuli should be presented in a different irregular order to each subject, but the first stimulus is usually chosen from among those in the middle region...." ([[12], p. 428], emphasis added).
Conclusion
A critical assumption of all studies using VAS-derived valuations is that the VAS is a proper interval scale, namely, the passage from 2 to 4 (2 points), for example, bears the same cardinal meaning as the passage from 56 to 58, and from 98 to 100 (as with a thermometer), with 0 and 100 arbitrarily chosen as reference points. If that assumption holds true, the analysis in this paper showed that the VAS valuation scores represent a value function as depicted in figure 4, with actual reference point at qp (valued at 50), and not a straight line diagonal connecting 100 (HRQL of perfect health) and 0 (HRQL of death).
The implications for HRQL measurement are that the verbal valuation is done in a relative way, with regard to a reference level of HRQL valued at 50. The exact level of HRQL, which is valued as 50, is unknown, and may vary across individuals. If it does vary across individuals, the comparison of VAS scores between individuals is problematic, since though the 0 and 100 anchors are well defined, they are actually used by the respondents to define the effective reference point qp evaluated as 50. Naturally, it does not mean that the S-shaped VAS score over- or under-estimate true HRQL relative to the common interpretation of VAS, since true HRQL is unknown. It does exclude, however, the notion of a reference point being the mean score in the population. The end-digit properties of written VAS evaluations done with the aid of a marked ruler are expected to be similar.
A straightforward test of the argument advanced in this paper would be to examine the distribution of VAS evaluations of own HRQL with respect to scores ending with 0, 5 and other integer by score-ranges in other populations, in particular where the traditional educational achievement scales are based on other scales, e.g., the A, B, C,...F grading system.
Acknowledgements
The research was partly funded by a grant from the National Institute for Health Policy Research in Israel. The comments of Zvi Adar on an earlier draft were very helpful.
==== Refs
Torrance GW Social preferences for health states: an empirical evaluation of three measurement techniques Socio-Economic Planning Sci 1978 10 129 138 10.1016/0038-0121(76)90036-7
Bleichrodt H Johannesson M An experimental test of a theoretical foundation for rating scale valuations Med Decis Making 1997 17 208 216 9107617
Lundberg L Johannesson M Isacson DGL Borgquist L Health state utilities in a general population in relation to age, gender and socioeconomic factors Eur J Pub Health 1999 9 211 217 10.1093/eurpub/9.3.211
Shmueli A Subjective health status and health values in the general population Med Decis Making 1999 19 122 127 10231074
EuroQol – A new facility for the measurement of health-related quality of life. The EuroQol Group Health Policy 1990 16 199 208 10109801 10.1016/0168-8510(90)90421-9
Robinson A Loomes G Jones-Lee M Visual analog scales, standard gambling and relative risk aversion Med Decis Making 2001 21 17 27 11206943
Torrance GW Feeny D Furlong W Visual Analog Scales: Do they have a role in the measurement of preferences for health states? Med Decis Making 2001 21 329 334 11475389 10.1177/02729890122062622
Denic S Khatib F Saadi H Quality of age data in patients from developing countries J Public Health 2004 26 168 171 10.1093/pubmed/fdh131
De Lusignan S Belsley J Hague N Dzregah B End-digit preference in blood pressure recordings of patients with ischemic heart disease in primary care J Hum Hypertens 2004 18 261 265 15037875 10.1038/sj.jhh.1001663
Shmueli A Israelis evaluate their health care system before and after the introduction of the National Health Insurance Law Health Policy 2003 63 279 287 12595127 10.1016/S0168-8510(02)00122-7
Kahnemann D Tversky A Prospect theory: an analysis of decision making under risk Econometrica 1979 47 263 291
Stevens SS Issues in psychophysical measurement Psychological Review 1971 78 426 450
|
16285884
|
PMC1308843
|
CC BY
|
2021-01-04 16:38:13
|
no
|
Health Qual Life Outcomes. 2005 Nov 14; 3:71
|
utf-8
|
Health Qual Life Outcomes
| 2,005 |
10.1186/1477-7525-3-71
|
oa_comm
|
==== Front
Health Qual Life OutcomesHealth and Quality of Life Outcomes1477-7525BioMed Central London 1477-7525-3-751630574810.1186/1477-7525-3-75ResearchDo proxies reflect patients' health concerns about urinary incontinence and gait problems? Higashi Takahiro [email protected] Ron D [email protected] Julie A [email protected] Caren J [email protected] Chau [email protected] David B [email protected] Paul G [email protected] David H [email protected] Roy T [email protected] Carol P [email protected] John T [email protected] Catherine H [email protected] Neil S [email protected] Department of Epidemiology and Healthcare Research, Kyoto University, Yoshida-Konoe-Cho Sakyo-ku, Kyoto, 606-8501, Japan2 UCLA Division of General Internal Medicine and Health Services Research: 911 Broxton Plaza 3rd Floor, Los Angeles, CA, 90095, USA3 RAND Santa Monica: 1776 Main Street P.O. Box 2138, Santa Monica, CA, 90407-2138, USA4 RAND Washington D.C.: 1200 South Hayes Street, Arlington VA 22202-5050, USA5 UCLA Division of Geriatrics: 10945 Le Conte Avenue Suite 2339, Los Angeles, CA 90095, USA6 Greater Los Angeles VA Healthcare System: 11301 Wilshire Boulevard, Los Angeles, CA 90073, USA2005 23 11 2005 3 75 75 13 9 2005 23 11 2005 Copyright © 2005 Higashi et al; licensee BioMed Central Ltd.2005Higashi et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
While falls and urinary incontinence are prevalent among older patients, who sometimes rely on proxies to provide their health information, the validity of proxy reports of concern about falls and urinary incontinence remains unknown.
Methods
Telephone interviews with 43 consecutive patients with falls or fear of falling and/or bothersome urinary incontinence and their proxies chosen by patients as most knowledgeable about their health. The questionnaire included items derived from the Medical Outcomes Study Short Form 12 (SF-12), a scale assessing concerns about urinary incontinence (UI), and a measure of fear of falling, the Falls Efficacy Scale (FES). Scores were estimated using items asking the proxy perspective (6 items from the SF-12, 10 items from a UI scale, and all 10 FES items). Proxy and patient scores were compared using intraclass correlation coefficients (ICC, one-way model). Variables associated with absolute agreement between patients and proxies were explored.
Results
Patients had a mean age of 81 years (range 75–93) and 67% were female while proxies had a mean age of 70 (range 42–87) and 49% were female. ICCs were 0.63 for the SF-12, 0.52 for the UI scale, and 0.29 for the FES. Proxies tended to understate patients' general health and incontinence concern, but overstate patients' concern about falling. Proxies who lived with patients and those who more often see patients more closely reflected patient FES scores compared to those who lived apart or those who saw patients less often. Internal consistency reliability of proxy responses was 0.62 for the SF-12, 0.86 for the I-QOL, and 0.93 for the FES. In addition, construct validity of the proxy FES scale was supported by greater proxy-perceived fear of falling for patients who received medical care after a fall during the past 12 months (p < .05).
Conclusion
Caution should be exercised when using proxies as a source of information about older patients' health perceptions. Questions asking about proxies' views yield suboptimal agreement with patient responses. However, proxy scales of UI and fall concern are internally consistent and may provide valid independent information.
Fear of fallingUrinary incontinenceHealth-related quality of lifePatient-proxy agreement
==== Body
Background
In addition to traditional objective measures of morbidity and mortality, self-reports are increasingly used to characterize patients' health and as an outcome of medical therapy. However, patients sometimes are unable to provide information because of cognitive impairment or other communication disabilities (e.g., hearing problems and/or language incompatibilities), or severity of illness. In such cases, investigators must decide whether to substitute the missing information with a proxy responder. This decision depends, at least in part, on the validity of proxy reports, usually conceptualized as how accurately the proxy reflects the information that would have been provided by the patient. Particularly in older populations, which are more likely to have cognitive impairment, studies have addressed this issue by comparing information provided independently by patients and proxies [1-9]. In general, these studies show good agreement between patients and proxies concerning observable behavior (such as physical function), but levels of concordance tend to be lower for internal perceptions such as energy level or emotional well-being [10].
Urinary incontinence and gait problems are prevalent among older persons. Studies show that up to one third of older individuals have at least occasional urinary incontinence [11]. Persons with urinary incontinence are twice as likely to report feeling depressed as their continent counterparts [12]. Similarly, nearly one-third of community-dwelling older persons fall each year, and up to half report a "fear of falling" [13,14]. Falls may lead to serious injury such as hip fracture, and fear of falling is associated with worse mental health and physical function [14,15].
Whether proxy information can be used to estimate fear of falling and incontinence in older patients is important for research and to improve care for these conditions. Measurement of intervention effects using self-report data would lead to the exclusion of substantial proportions of patients with these conditions who are unable to provide these data. On the other hand, noise or bias is introduced into measurement if proxy responses do not accurately reflect the perspective of patients. As part of a quality improvement intervention focused on care for falls and incontinence for older patients, we evaluated the validity of proxy responses assessing patient perceptions of health, fear of falling and incontinence. We also identified variables that were significantly associated with patient-proxy agreement.
Methods
Sample
This study examined proxy-reported measures that could supplement patient reports in the evaluation of an intervention to improve the quality of outpatient care for urinary incontinence, falls and gait impairment, and cognitive impairment. Details of this controlled trial are described elsewhere [16]. Data presented in this report are from the enrollment phase, before any intervention. Consecutive community-dwelling patients age 75 years or older receiving care from two medical groups in southern California were interviewed by telephone to screen for these three conditions several days before a scheduled visit to their physician. Patients identified as having any of the three conditions were invited to participate in the practice-based quality improvement project.
Among 649 patients who consented to participate, 531 patients answered questions for themselves, and 118 proxies provided information for patients who could not provide information. For the purpose of examining the concordance in response between patients and proxies, we asked 44 consecutive patients among the 531 patients who provided self-report information (25 had urinary incontinence, 32 had falls and/or fear of falling) to name a "proxy," defined as the person most knowledgeable about his/her health. Proxies were contacted to participate in a telephone interview. Although proxies were interviewed separately, the majority of proxy interviews were conducted on the same day as the patient interview. The RAND and University of California, Los Angeles Institutional Review Boards approved the study protocol (UCLA IRB#G02-03-002, RAND-IRB#00000051).
Measurements
We collected information via telephone interview on the SF-12 Health Survey [17], the Urinary Incontinence Quality of Life (I-QOL) scale [18], the Falls Efficacy Scale (FES) [14], and a proxy urinary incontinence (pUI) survey (described below). Briefly, the SF-12 is an abridged version of the Short Form-36 Health Survey [17,19,20] that measures 8 domains of health. Although the SF-12 is only one third the length of the SF-36, it accounts for more than 90% of the variance in the SF-36 physical and mental health summary scores in the U.S. population [17]. The SF-12 physical component and mental component scores (summary measures) are standardized to the U.S. normative population with a mean of 50 and standard deviation of 10. A higher score means better health. The I-QOL scale assesses patients' concerns about urinary incontinence using 22 items covering 3 domains of concern about urinary incontinence: "avoidance and limiting behavior," "psychosocial impacts," and "social embarrassment." Each item specifies patients' concern or limitation of activities due to urinary incontinence. Patients rate these items with 5 response options from "extremely" true to "not at all" true. The score of the I-QOL is the sum of the item responses converted to a 0–100 possible range with a higher score signifying higher quality of life. The FES assesses fear of falling during daily activities in older persons [14]. It consists of 10 items with 4 possible responses ranging from "not concerned at all" to "very concerned." Scores are computed as the sum of item responses, ranging from 10 to 40. Although the original scale was created so that a larger number indicated greater concern, we reversed the scale to make it consistent with the SF-12 and I-QOL scales. Therefore, a higher FES score signifies less concern about falling.
Because the aim of this study was to compare patient and proxy responses, we conducted separate telephone interviews with patients and proxies. After examination of the feasibility of administration, we selected for proxy administration a subset of items from the SF-12 (4 items) and created a new proxy urinary incontinence (pUI) scale from 10 items modified from the I-QOL scale. Eight items from the SF-12 and 12 items from the I-QOL scales were excluded because they were judged to measure internal perceptions, for which the literature shows poor agreement between proxy and patient responses [10], and we predicted that it would be difficult to obtain a proxy's view on these issues. Two additional physical function items from the SF-36 (capability of vigorous activities and difficulty in bathing or dressing) not contained in the SF-12 also were asked in both patient and proxy interviews. The items contained in the proxy interview are listed in Table 1. All items in the FES were judged as feasible for proxy interview. For survey items that assessed a patient's concern about urinary incontinence and falling, we asked the proxy's concern about patients rather than querying the proxy's opinion of the patient's concern to avoid confusing responding proxies.
Table 1 Item Descriptor, Number of Patient-Proxy Pairs and Intraclass-Correlation Coefficients of Individual Items in the pSF-12p, pUI Scale, and FES
Item descriptor N ICC* %Bias†
pSF-12 PCS Items
Self-report health (poor/fair/good/very good/excellent) 41 0.38 8.5%
Limitation in vigorous activities? 43 0.27 3.5%
Limitation moderate activities? 42 0.54 9.5%
Limitation in climbing several flights of stairs 43 0.39 8.1%
Limitation in dressing/bathing yourself 43 0.08 9.3%
Limited kind of activities/work 43 0.09 -16.3%
pUI Scale Items‡
Worry about getting to the toilet on time 24 0.64 -8.3%
Have to be careful about sitting/standing 23 0.60 -1.1%
Worry where the toilets are in new places 24 0.51 0.0%
Don't feel free to leave home 24 0.14 -3.1%
Worry about others smelling urine on me 24 0.39 -9.4%
Frequent trip to toilet is important 24 0.43 -5.2%
Plan details in advance 24 0.29 -8.3%
Hard to get good sleep 24 0.63 -4.2%
Watch what/how much to drink 24 0.07 -4.2%
Limited choice of clothing 22 -0.10 1.1%
FES Items‡
Cleaning the house 32 0.10 17.7%
Getting dressed 32 0.23 9.4%
Preparing simple meals 32 0.38 -3.1%
Taking a bath/shower 32 0.26 3.1%
Simple shopping 30 0.10 15.6%
Getting in and out of a chair 32 0.23 6.3%
Going up/down stairs 32 0.02 24.0%
Walking around the neighborhood 32 0.10 18.8%
Reaching into cabinets or closets 32 0.19 0.0%
Going to answer the telephone 32 0.09 8.3%
* Intraclass correlation
† * Mean proxy answers compared to patient answers on the item expressed as the percentage of the full scale points. Negative values indicate proxies indicated worse health on pSF and less concerned on FES and pUI.
‡ For proxy interview, the pUI and FES items were modified to address proxy's concerns about patient UI and fear of patient falling
Since only subsets of items were used in proxy interviews for the SF-12 PCS and I-QOL scales, we predicted full-item scores for these scales (termed pSF-12 PCS and pUI, respectively) using a weighted combination of items. Weights for items in the pSF-12 PCS and pUI were obtained from the coefficients from linear regression models with dependent variables of SF-12/I-QOL scores calculated from the full set of items to which patient responded, and the predictor variables of patient responses to items included in proxy interviews (6 items for SF-12 and 10 items for I-QOL). Coefficients used to compute pSF-12 PCS and pUI scales are available from the first author. The regression models were fitted using responses from the 531 interviewed patients. For the regression, we performed a complete case analysis excluding cases with any missing items (SF-12: N = 488, I-QOL: N = 179). The estimated SF-12 PCS and I-QOL scores accounted for 85% and 92% of the variance, respectively, of the scores calculated by the full set of items. We did not calculate SF-12 mental component scores because published information indicates that proxies do not provide valid information about patients' mental health.
Five proxies did not answer one or two items (2 proxies for pSF-12, 1 for pUI scale, 1 for the FES and 1 for both the pUI scale and the FES). We computed alternative coefficients for the available items based on regression models without the missing variables. One patient failed to answer one item of the pSF-12 PCS, so the score for this patient was estimated from a separate regression model using 11 available items and 2 extra SF items. For the I-QOL scale and the FES, missing items were imputed from the mean of the other item responses when the number of missing items was 3 or fewer for I-QOL and 2 or fewer for FES. One proxy was excluded from the analysis because he did not answer more than half of the scale items. For the purpose of comparison, the predicted scores from the regression model (i.e., pSF-12, pUI scores rather than original SF-12, or I-QOL scores) were used for both patient and proxy.
Additionally, a self-administered survey was mailed to proxies asking about their own age, education level, whether or not they lived with the patient, how often they saw the patient, how often they accompanied the patient to physician visits, and their level of knowledge about various aspects of the patient's health, medication, mood and activities.
Statistical analysis
Proxy scores on the pSF-12 PCS, pUI, and FES scales were compared with patient response scores using intraclass correlation coefficients (ICC, one-way model) that represent both correlations and systematic mean differences [21]. In order to evaluate the relationship between proxy characteristics and the agreement of proxy-patient responses, the absolute difference in scores was computed. The absolute difference represents the discrepancy between the proxy and patient reponses ignoring the direction of the difference. Since score ranges varied across scales, the absolute difference was also expressed in terms of the patient score standard deviation as well as the percentage of each scale range. We used one-way ANOVA to evaluate mean absolute difference in scores across groups of proxies if the assumption of equal variance held; otherwise the Kruskal-Wallis test was used. The assumption of equal variance was considered to be violated when Bartlett's test rejected the null hypotheses at the level of p = 0.10. Otherwise, the statistical signficance level was set at 0.050.
We used the full 118 proxies from the intervention study for the purpose of examining internal consistency reliability of the pSF-12 PCS, pUI and the FESusing Cronbach's alpha [22]. For the calculation of alpha, item responses were assigned the weights (available from the first author) used in the score calculation for the pSF-12 PCS and pUI. In addition, construct validity of the proxy response FES scale was assessed. For FES, the proxies for patients were divided into 3 groups of ascending fall severity: 1) those who had not fallen in the past 12 months, 2) those who had fallen but did not need to see a provider, and 3) those who had fallen and needed to see a provider. Scores across these groups were tested using the Cuzick's test for trend [23]. Statistical analysis was performed using STATA version 8.2 (STATA Corporation, College Station, TX).
Results
Patient and proxy characteristics
The 43 patients in this analysis had a mean age of 81 (range 75–93) and 29 (67%) were female. Proxies had a mean age of 70 (range 42–87) and 21 (49%) were female. Proxies were related to patients as follows: 16 husbands (37%), 11 daughters (26%), 7 wives (16%), 4 sons (9%), and 5 others (2 roommates, 1 sister, 1 fiancé and 1 niece). Thirty-five proxies lived with the patients. One of the 43 proxies who completed a telephone interview did not return the mail survey and was excluded from the analysis of the association between proxy characteristics and agreement. A summary of the scores for the SF-12 PCS, FES, pUI and pSF-12 PCS for the patients and proxies in this study, as well as all patients and proxies in the intervention study sample, are presented in Table 2. The scores on these scales for the 43 patients in this substudy were not different from the full patient intervention study sample.
Table 2 Mean Scores for Study Sample and the Full Intervention Project Patient Sample
Dyad study sample
Patient Proxy
Mean SD# N Mean SD# N
SF-12 PCS* 37.4 (11.5) 43
pSF-12 PCS† 37.0 (9.9) 43 36.2 (9.2) 43
I-QOL‡ 78.1 (17.8) 24
pUI Scale§ 76.5 (18.4) 24 80.9 (20.0) 24
FES¶ 31.8 (7.2) 32 28.8 (7.0) 32
Full intervention project sample
Patient Proxy
Mean SD# N Mean SD# N
SF-12 PCS* 36.7 (10.7) 492
pSF-12 PCS† 36.6 (9.8) 509 31.8 (9.7) 117
I-QOL‡ 73.4 (20.4) 203
pUI Scale§ 73.4 (19.6) 198 50.7 (23.6) 26
FES¶ 30.4 (7.8) 411 21.6 (9.2) 72
*SF-12 PCS: Short Form 12 physical component score †pSF-12: proxy version of SF-12
‡I-QOL: Incontinence quality of life scale §pUI: proxy urinary incontinence, ¶FES: falls efficacy scale, #SD: standard deviation.
Note: Number of patients that had SF-12 and pSF-12 scales are different because some patients had missing values.
Agreement between patients and proxies
Intraclass correlations between patients and proxies on the predicted scores were 0.63 for pSF-12 PCS, 0.52 for pUI scores, and 0.29 for the FES scores. The distribution of the differences between patient scores and proxy scores are illustrated in Figure 1. On average, proxy scores were lower than patient scores by 0.8 for the pSF-12 PCS (i.e., proxies underestimate patients' health status), 3.0 points lower on the 40-point FES (i.e., proxies are more concerned than patients about patients falling), and 4.4 higher on the 0–100 pUI scores (i.e., proxies are less concerned about incontinence than patients). The mean absolute difference in scores between patients and proxies was 6.2 for the pSF-12 (0.6 standard deviation (SD)), 13.0 for the pUI (0.7 SD) and 6.2 for the FES (0.9 SD). Ninety percent of proxy scores fell within 14 absolute difference points for pSF-12 scores (1.4 SD), 32 points for pUI scores (1.7 SD), and 15 points for FES scores (2.1 SD). One proxy, a patient's fiancé, provided answers to FES items that were extremely discrepant from the patient's answers, resulting in an FES score difference of 23. Excluding this outlier pair, the ICC, mean raw and absolute difference between patient score and proxy score (patient score minus proxy score) for the FES was 0.40, 3.8 (i.e., proxies are more concerned), and 5.6 (19% of possible score range, 0.8 SD), respectively.
Table 3 shows the relationship of the mean absolute difference of each scale to proxy characteristics. Proxy characteristics were generally unrelated to agreement with patient reported health. However, proxies who lived with patients and those who saw patients every day had a significantly smaller mean absolute difference from patient scores on the FES than those who lived apart or those who saw patients less frequently. Proxy reports of familiarity with patients' medications, activity and mood were unrelated to agreement with patient reports on any scale. Similarly, proxy educational level and relationship with the patient did not predict accuracy concerning physical health or incontinence or falls concerns.
Figure 1 Differences in estimated SF-12 PCS, pUI, and FES scores between proxies and patients (proxy score) – (patient score). * >0 indicates proxy score is greater than patient score.
Table 3 Mean Absolute Difference between Patient and Proxy Responses for pSF-12 Physical Component Score(PCS), pUI scale, and FES Scores
pSF-12 PCS pUI scale FES
Questions N mean absolute difference (P value) N mean absolute difference (P value) N mean absolute difference (P value)
Relationship of proxy to patient Husband 16 5.4 12 11.9 10 5.4
Wife 7 9.8 2 4.3 6 5.3
Son 4 5.8 2 30.2 4 0.9
Daughter 11 4.8 5 5.3 8 8.7
Others+ 5 7.0 (0.32) 3 24.7 (0.16)* 4 9.8 (0.14)*
Proxy's gender Male 21 5.7 14 14.5 15 4.1
Female 22 6.2 (0.52) 10 10.9 (0.52) 17 8.0 (0.06)*
Proxy's educational level Professional 7 7.0 4 10.7 4 8.0
College 14 5.5 8 16.6 13 7.4
Vocational 14 6.0 7 15.2 10 4.4
High School 7 7.2 (0.87) 4 7.6 (0.70) 5 5.2 (0.60)
Proxy lives with patient? Yes 35 6.6 21 11.5 25 4.8
No 8 4.4 (0.28) 3 24.0 (0.12) 7 11.0 (0.05)*
How often proxy accompanies patient to MD visit? Always 17 7.5 11 13.9 13 5.7
Usually 7 7.2 3 10.1 5 5.0
Sometimes 13 4.4 6 14.7 11 7.5
Never 5 5.0 (0.40) 3 13.8 (0.97) 3 5.7 (0.85)
Days per week proxy sees patient? 7 days 29 5.9 17 13.1 20 4.3
<7 days 9 4.7 (0.51) 3 24.0 (0.21) 8 10.1 (0.05)*
No answer 4 11.5 3 6.0 4 7.6
How well proxy knows about patient's health? Very well 29 6.5 15 11.1 21 7.7
Pretty well 10 6.1 6 15.3 9 3.8
Somewhat 3 3.2 (0.59) 2 27.0 (0.26) 2 1.2 (0.08)*
Is proxy familiar with patient's meds? Yes 36 6.3 19 13.9 27 6.6
No 3 4.1 2 3.6 3 1.0
Not sure 2 7.9 1 36.6 2 8.5
Not Taking 1 4.0 (0.83) 1 4.4 (0.19) 0 (0.25)
How well proxy knows about patient's activity? Very well 36 6.2 22 14.0 27 6.7
Pretty well 5 5.3 1 5.0 4 4.0
Somewhat 1 11.2 (0.59) 0 (0.52) 1 1.0 (0.48)
How well proxy knows about patient's mood? Very well 24 6.3 11 13.0 18 7.4
Pretty well 16 6.3 12 14.1 12 4.7
Somewhat 2 3.8 (0.81) 0 (0.84) 2 4.0 (0.42)
Note: Information on relationship to patient, gender and whether or not proxy lives with patient was available from patients, thus, includes a person who did not respond to the mail survey.
* P value derived from Kruskal-Wallis test because test for equal variance did not hold.
+ Others included roommate fiancé, niece, and sister.
Intraclass correlation coefficients of individual items ranged from -0.10 to 0.64 (Table 1). The items with poorest agreement among pSF-12 items were whether "health limits dressing/bathing" and whether "health limits kind of activities/work,"; among pUI items, whether "urinary incontinence limits choice of clothing" and whether he/she has to "watch what/how much to drink"; and among FES items, whether there is concern about falling when "going up/down stairs," "going to answer the telephone," doing "simple shopping," and "walking around the neighborhood". The best agreement was found for whether "health limits moderate activities" among pSF-12 items, "worry about getting to the toilet on time" among pUI items and concern about falling "when preparing simple meals" among FES items.
Reliability and validity of proxy responses
Cronbach's alphas for proxy response scales were 0.62 for the pSF-12 PCS scale, 0.86 for pUI scale and 0.93 for the FES. Table 4 shows the correlation coefficients between individual items and the scale, and scale alphas when the score was calculated without each item.
Table 4 Correlation between Item and Scale and Cronbach's Alpha of Proxy Scales When Each Item is Deleted
N of proxies Item-scale correlation Cronbach's alpha*
pSF-12 PCS Items
Self-report health (poor/fair/good/very good/excellent) 118 0.57 0.58
Limitation in vigorous activities? 118 0.55 0.62
Limitation in moderate activities? 118 0.81 0.47
Limitation in climbing several flights of stairs 118 0.63 0.55
Limitation in dressing yourself 118 0.26 0.64
Limited kind of activities/work 117 0.79 0.54
Overall = 0.62
pUI Scale Items
Worry about getting to toilet on time 28 0.72 0.84
Have to be careful about sitting/standing 27 0.77 0.84
Worry where the toilets are in new places 28 0.72 0.84
Don't feel free to leave home 27 0.69 0.84
Worry about others smelling urine on me 28 0.84 0.83
Frequent trip to toilet is important 28 0.67 0.85
Plan details in advance 28 0.70 0.85
Hard to get good sleep 27 0.43 0.87
Watch what/how much to drink 28 0.54 0.86
Limited choice of clothing 28 0.72 0.85
Overall = 0.86
FES Items
Cleaning the house 69 0.79 0.93
Getting dressed 81 0.83 0.92
Preparing simple meals 68 0.84 0.92
Taking a bath/shower 79 0.83 0.93
Simple shopping 62 0.84 0.93
Getting in and out of a chair 81 0.79 0.93
Going up/down stairs 74 0.69 0.93
Walking around the neighborhood 72 0.82 0.93
Reaching into cabinets or closets 76 0.75 0.93
Going to answer the telephone 71 0.77 0.93
Overall = 0.93
* Cronbach's alpha of scales when each item is delete
The mean FES proxy score was 26.0 (n = 20) for the patients who hadn't fallen in the past 12 months, 21.8 for the patients who had fallen but didn't need to see a provider (n = 22), and 18.4 for patients who had fallen and needed to see a provider (n = 18). This score trend was statistically significant (p = 0.009).
Discussion
This study demonstrated that agreement between proxies and patients is not very high for reports about physical health and concerns about incontinence and falls. Neither the pSF-12 PCS, the pUI nor the FES reached the intraclass correlation coefficient cut off of 0.7 that is generally considered acceptable agreement [24]. Directionality of difference in scores indicated that proxies tended to underestimate patient health and be more concerned about patient gait problems, but less concerned about urinary incontinence. Compared to prior literature, which generally shows that proxies overestimate functional impairment, our findings for the SF-12 and the FES are consistent, but the finding for the concern about urinary incontinence is not. This may reflect privacy concerns regarding urinary incontinence and patients' reluctance to talk about it. Alternatively, it may suggest that fear of falling is of greater concern because of the potential for injury, disability and financial costs.
Despite suboptimal agreement with patient responses, proxy responses appeared to be internally consistent and the results of this study provide support for construct validity of proxy FES, suggesting that these scales may measure proxy responses that represent important constructs that are distinct from the patient responses. Our decision to ask proxies' concern rather than proxy estimate of patients' concern may have strengthened this tendency because proxy's own concern will be less influenced by guessing. The alternative approach of asking the proxy to estimate the patient's level of concern may be more likely to produce missing values and becomes particularly problematic for patients with severely compromised cognition. However, further study should test proxy items that ask about patient concern to evaluate whether these questions better reflect patient responses.
Proxies' individual characteristics did not predict agreement between patients and proxies on these three scales. Though most proxies reported that they were knowledgeable about patients' health, mood, activities and medications, they did not well replicate patient responses, and proxy report of familiarity with these issues was unrelated to agreement. Rather than self-assessment of familiarity, structural characteristics such as whether the patient and proxy live together or how often they see each other were better associated with proxy agreement with patient scores, at least for the FES.
In our study, the comparison across proxy characteristics was made only among proxies who were considered most knowledgeable about the patients' health. A subject could have identified a proxy who was most knowledgeable about his/her health but did not live with him/her. Our finding of better prediction of agreement by structural characteristics does not necessarily mean that selecting proxies primarily based on structural characteristics will result in better proxies. Future study is necessary to explore the best strategy to select optimal proxies.
Our study has several limitations. First, patients in the analyses of agreement had intact cognitive function and were well enough to participate in the survey process so their proxies may not have needed to be familiar with patient health status and concerns, since patients could communicate this information on their own. "Real" proxies who take care of patients may provide more accurate information than the "forced" proxies used for practical reasons in this study, though this is impossible to prove. Second, we used only items from the SF-12 and I-QOL scales that appeared to be feasible for proxy interview and we approximated the proxy score by re-weighting these responses. Since we included only items that we judged to be feasible, this may overestimate the agreement compared to the entire set of items from the original scales. Finally, the sample size of this study was small and there was limited statistical power for the evaluation of proxy factors associated with agreement. Nonetheless, our findings are generally consistent with prior research regarding the effect of proxy characteristics on patient-proxy agreement: demographics such as age, gender and education are inconsistent predictors of agreement [25-29], while greater contact such as living together or frequent visits are reasonable, though not perfect, predictors of agreement [9,25,29]. Larger studies will need to identify the role of factors related to the accuracy of proxy health reports for patient subjective health.
Conclusion
Our study shows that the agreement between patient and proxy responses for SF-12, urinary incontinence and FES scales are less than acceptable. However, there is support for the reliability and construct validity of proxy responses. Rather than replacing patient responses, proxy responses may be better used in separate analysis. Future research should evaluate other scale properties, including test-retest reliability and responsiveness to change, of proxy responses, which may, on their own, be valuable scales.
Authors' contributions
TH, RH, DR, PG, CR, and NW contributed to concept and design, acquisition, analysis, and interpretation of data and preparation of manuscript. JB, CJ, and CP contributed to concept and design, acquisition of data and preparation of manuscript. DS and CM contributed to concept and design, analysis and interpretation of data and preparation of manuscript. RY contributed to concept and design, interpretation of data and preparation of manuscript. JC contributed to concept and design, and interpretation of data and preparation of manuscript. All authors read and approved the final manuscript.
Acknowledgements
Robin P. Hertz, PhD, senior director of outcomes research/population studies at Pfizer Inc, provided valuable support. We recognize the technical assistance of Patricia Smith. This project was supported by a contract from Pfizer Inc to RAND. The sponsor had no role in the design, methods, subject recruitment, data collections, analysis and preparation of paper.
Part of the work was presented in the 26th National Meeting of Society of General Internal Medicine.
Dr. Higashi was supported by a St. Luke's Life Science Institute Fellowship Award. Dr. Hays was supported by the UCLA/DREW Project EXPORT, National Institutes of Health, National Center on Minority Health & Health Disparities, (P20-MD00148-01) and the UCLA Center for Health Improvement in Minority Elders/Resource Centers for Minority Aging Research, National Institutes of Health, National Institute of Aging, (AG-02-004). Dr. Shekelle is a Senior Research Associate of the Veterans Affairs Health Services Research & Development Service. Dr. Chang is supported by a National Research Service Award (PE-19001) and the UCLA Specialty Training and Advanced Research (STAR) Program. Dr. MacLean is a recipient of a Veterans Affairs Health Services Research & Development Service Career Development Award.
==== Refs
Yip JY Wilber KH Myrtle RC Grazman DN Comparison of older adult subject and proxy responses on the SF-36 health-related quality of life instrument Aging Ment Health 2001 5 136 142 11511060 10.1080/13607860120038357
Magaziner J Zimmerman SI Gruber-Baldini AL Hebel JR Fox KM Proxy reporting in five areas of functional status. Comparison with self-reports and observations of performance Am J Epidemiol 1997 146 418 428 9290502
Yasuda N Zimmerman S Hawkes WG Gruber-Baldini AL Hebel JR Magaziner J Concordance of proxy-perceived change and measured change in multiple domains of function in older persons J Am Geriatr Soc 2004 52 1157 1162 15209655 10.1111/j.1532-5415.2004.52315.x
Ball AE Russell EM Seymour DG Primrose WR Garratt AM Problems in using health survey questionnaires in older patients with physical disabilities. Can proxies be used to complete the SF-36? Gerontology 2001 47 334 340 11721148 10.1159/000052824
Pierre U Wood-Dauphinee S Korner-Bitensky N Gayton D Hanley J Proxy use of the Canadian SF-36 in rating health status of the disabled elderly J Clin Epidemiol 1998 51 983 990 9817116 10.1016/S0895-4356(98)00090-0
Andresen EM Vahle VJ Lollar D Proxy reliability: health-related quality of life (HRQoL) measures for people with disability Qual Life Res 2001 10 609 619 11822794 10.1023/A:1013187903591
Rubenstein LZ Schairer C Wieland GD Kane R Systematic biases in functional status assessment of elderly adults: effects of different data sources J Gerontol 1984 39 686 691 6436360
Magaziner J Simonsick EM Kashner TM Hebel JR Patient-proxy response comparability on measures of patient health and functional status J Clin Epidemiol 1988 41 1065 1074 3204417 10.1016/0895-4356(88)90076-5
Bassett SS Magaziner J Hebel JR Reliability of proxy response on mental health indices for aged, community-dwelling women Psychol Aging 1990 5 127 132 2317291 10.1037/0882-7974.5.1.127
Sneeuw KC Sprangers MA Aaronson NK The role of health care providers and significant others in evaluating the quality of life of patients with chronic disease J Clin Epidemiol 2002 55 1130 1143 12507678 10.1016/S0895-4356(02)00479-1
Herzog AR Fultz NH Prevalence and incidence of urinary incontinence in community-dwelling populations J Am Geriatr Soc 1990 38 273 281 2179368
Dugan E Cohen SJ Bland DR Preisser JS Davis CC Suggs PK McGann P The association of depressive symptoms and urinary incontinence among older adults J Am Geriatr Soc 2000 48 413 416 10798468
Walker JE Howland J Falls and fear of falling among elderly persons living in the community: occupational therapy interventions Am J Occup Ther 1991 45 119 122 2035588
Tinetti ME Mendes de Leon CF Doucette JT Baker DI Fear of falling and fall-related efficacy in relationship to functioning among community-living elders J Gerontol 1994 49 M140 7 8169336
Kressig RW Wolf SL Sattin RW O'Grady M Greenspan A Curns A Kutner M Associations of demographic, functional, and behavioral characteristics with activity-related fear of falling among older adults transitioning to frailty J Am Geriatr Soc 2001 49 1456 1462 11890583 10.1046/j.1532-5415.2001.4911237.x
Reuben DB Roth C Kamberg C Wenger NS Restructuring primary care practices to manage geriatric syndromes: the ACOVE-2 intervention J Am Geriatr Soc 2003 51 1787 1793 14687359 10.1046/j.1532-5415.2003.51565.x
Ware JJ Kosinski M Keller SD A 12-Item Short-Form Health Survey: construction of scales and preliminary tests of reliability and validity Med Care 1996 34 220 233 8628042 10.1097/00005650-199603000-00003
Wagner TH Patrick DL Bavendam TG Martin ML Buesching DP Quality of life of persons with urinary incontinence: development of a new measure Urology 1996 47 67 72 8560665 10.1016/S0090-4295(99)80384-7
McHorney CA Ware JEJ Lu JF Sherbourne CD The MOS 36-item Short-Form Health Survey (SF-36): III. Tests of data quality, scaling assumptions, and reliability across diverse patient groups Med Care 1994 32 40 66 8277801
McHorney CA Ware JEJ Raczek AE The MOS 36-Item Short-Form Health Survey (SF-36): II. Psychometric and clinical tests of validity in measuring physical and mental health constructs Med Care 1993 31 247 263 8450681
Winer BJ Brown DR Michels KM Statistical Principles in Experimental Design McGraw-Hill Series in Psychology 1991 3rd New York , McGraw-Hill
Cronbach LJ Coefficient alpha and the internal structure of tests. . Psychometrika 1951 16 297 334
Cuzick J A Wilcoxon-type test for trend Stat Med 1985 4 87 90 3992076
Nunnally JC Psychometric theory 1978 2nd New York , McGraw-Hill
McCusker J Stoddard AM Use of a surrogate for the Sickness Impact Profile Med Care 1984 22 789 795 6492907
Epstein AM Hall JA Tognetti J Son LH Conant LJ Using proxies to evaluate quality of life. Can they provide valid information about patients' health status and satisfaction with medical care? Med Care 1989 27 S91 8 2921889
Hays RD Vickrey BG Hermann BP Perrine K Cramer J Meador K Spritzer K Devinsky O Agreement between self reports and proxy reports of quality of life in epilepsy patients Qual Life Res 1995 4 159 168 7780382 10.1007/BF01833609
Rothman ML Hedrick SC Bulcroft KA Hickam DH Rubenstein LZ The validity of proxy-generated scores as measures of patient health status Med Care 1991 29 115 124 1994145
Magaziner J Bassett SS Hebel JR Gruber-Baldini A Use of proxies to measure health and functional status in epidemiologic studies of community-dwelling women aged 65 years and older Am J Epidemiol 1996 143 283 292 8561163
|
16305748
|
PMC1308844
|
CC BY
|
2021-01-04 16:38:13
|
no
|
Health Qual Life Outcomes. 2005 Nov 23; 3:75
|
utf-8
|
Health Qual Life Outcomes
| 2,005 |
10.1186/1477-7525-3-75
|
oa_comm
|
==== Front
Int J Health GeogrInternational Journal of Health Geographics1476-072XBioMed Central London 1476-072X-4-271626907810.1186/1476-072X-4-27ResearchTowards understanding the presence/absence of Human African Trypanosomosis in a focus of Côte d'Ivoire: a spatial analysis of the pathogenic system Courtin Fabrice [email protected] Vincent [email protected]é Emmanuel [email protected] Bamoro [email protected] Yohan [email protected] Sophie [email protected] Gérard [email protected] Jean-Pierre [email protected] Philippe [email protected] Institut Pierre Richet (IPR), équipe « THA et glossines », s/c IRD, Rue Fleming zone 4C, 04 BP 293, Abidjan 04, Côte d'Ivoire2 Institut de Recherche pour le Développement (IRD), Unité de Recherche UR 177, Laboratoire de Recherche et de Coordination sur les Trypanosomoses (LRCT IRD-CIRAD), TA 207/G, Campus International de Baillarguet, 34398 Montpellier cedex 5, France3 University of Lille, USTL/LGMA (Laboratoire de Géographie des Milieux Anthropisés), UMR CNRS 8141, France4 University of Montpellier 3, UFR sciences humaines et sciences de l'environnement, route de Mende 34199 Montpellier Cedex 5, Laboratoire de recherche GESTER (gestion des territoires et des risques), France2005 3 11 2005 4 27 27 25 8 2005 3 11 2005 Copyright © 2005 Courtin et al; licensee BioMed Central Ltd.2005Courtin et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
This study aimed at identifying factors influencing the development of Human African Trypanosomosis (HAT, or sleeping sickness) in the focus of Bonon, located in the mesophile forest of Côte d'Ivoire. A previous study mapping the main daytime activity sites of 96 patients revealed an important disparity between the area south of the town- where all the patients lived- and the area north of the town, apparently free of disease. In order to explain this disparity, we carried out a spatial analysis of the key components of the pathogenic system, i.e. the human host, the tsetse vector and the trypanosomes in their environment using a geographic information system (GIS).
Results
This approach at the scale of a HAT focus enabled us to identify spatial patterns which linked to the transmission and the dissemination of this disease. The history of human settlement (with the rural northern area exploited much earlier than the southern one) appears to be a major factor which determines the land use pattern, which itself may account for differences found in vector densities (tsetse were found six times more abundant in the southern rural area than in the northern). Vector density, according to the human and environmental context in which it is found (here an intense mobility between the town of Bonon and the rural areas), may explain the observed spatial differences in HAT prevalence.
Conclusion
This work demonstrates the role of GIS analyses of key components of the pathogenic system in providing a better understanding of transmission and dissemination of HAT. Moreover, following the identification of the most active transmission areas, and of an area unfavourable to HAT transmission, this study more precisely delineates the boundaries of the Bonon focus. As a follow-up, targeted tsetse control activities starting north of Bonon (with few chances of reinvasion due to very low densities) going south, and additional medical surveys in the south will be proposed to the Ivoirian HAT control program to enhance the control of the disease in this focus. This work also shows the evolution of HAT regarding time and environment, and the methodology used may be able to predict possible sleeping sickness development/extinction in areas with similar history and space organization.
==== Body
Background
Human African Trypanosomosis (HAT) or sleeping sickness, is a vector-borne parasitic disease of Sub-Saharan Africa. The parasite (Trypanosoma brucei gambiense in West and Central Africa) is transmitted to humans by the bite of a dipteran insect, the tsetse fly (Glossina sp.). The disease, though almost eradicated in the early 60's, has once again become a major public health problem. Currently, about 60 million people are at risk of infection and around 300,000 are estimated to have the disease [1]. Two principal parameters are usually put forward to explain this resurgence. First, the routine measures that were implemented to control the disease have gradually disappeared [2]. Secondly, changes in the environment, through their effect on the relationships between host, vector and parasite, may also account for a significant part of the disease's re-emergence [3]. In the forest area of Côte d'Ivoire, sleeping sickness has usually been associated with coffee and cocoa plantations [4]. The establishment of these cash crop plantations, combined with massive immigration of agricultural labour, has altered the original habitat and caused the disappearance of the mainly zoophilic forest tsetse fly species, which have been replaced by vectors with a more opportunistic feeding pattern such as the main vector of sleeping sickness G. palpalis, that are able to adapt to peri-urban or urban areas [5-8]. The rise in numbers of agricultural workers also led to increased vector-host contact [9]. The establishment of new villages (defined as inhabited by several families with a chief) and more especially new encampments (defined as inhabited by one family or by agricultural labourers, in coffee/cocoa plantations) also increased levels of movement along the new communication routes, further increasing human-vector contact [10-12]. Though such change is now widely accepted as a major cause of the development of sleeping sickness in the forest area of West Africa, it has yet to be discovered why the disease is present in some places but not in others, where demographic, behavioural and environmental conditions appear to be similar.
The focus of Bonon is located in the mesophile forest of Côte d'Ivoire (figure 1). The evolution of HAT prevalence in this focus since the 1950s (data supplied by the HAT National Control Program), enables to trace back the appearance and the spatial evolution of disease. The disease has been endemic for a long time with 9 cases being detected between 1956 and 1959, 57 between 1976 and 1985, 102 between 1986 and 1996, then 16 in the year 1997. These patients came principally from the south rural area. Since then 3 medical surveys in 1998 and 1999 detected a further 33 cases from visits to 7,824 people also principally from the south rural area. The following year, an exhaustive medical survey was undertaken in the area during which 15 000 people were visited and 74 cases detected, none of these cases living in the north rural area. Previous work has mapped the places of residence, water supply sites and day time activity of 96 patients who were diagnosed to have sleeping sickness. This mapping revealed an important disparity between the rural area south of the town and the rural area north of the town: the southern rural area was very affected by HAT, as shown by a high prevalence, and a high rate of activity by urban patients. By contrast, the northern rural area was characterized by a low prevalence and by a very low attendance of urban patients (figure 1). This disparity allowed to identify different areas where the mechanisms of HAT transmission may differ [11]. The aim of the present work was thus to carry out a spatial analysis of key components of the pathogenic system, taking into account parasitological, entomological, human and environmental factors, in order to try to explain the appearance and the maintenance of HAT in the southern area of this focus and its absence in the northern area. The work was conducted in four steps: (1) a study of the relationship between tsetse fly distribution and the daily activity patterns of patients, (2) a study of the history of human settlement in the focus, including the relationships between the different ethnic groups and observed land use patterns, (3) a detailed description of the land cover and land use of the area using a ground transect approach, combined with remote sensing analyses, and (4) the characterization of spatial human mobility patterns, particularly between the town of Bonon and its surroundings.
Figure 1 (Solano et al. 2003; modified)- Geographic location of Bonon, daily routes and residence places of HAT patients. The upper maps show the locations of Côte d'Ivoire in Africa and the focus of Bonon in Côte d'Ivoire. The bottom map shows the living places and the mobility of patients. The majority of patients live in or move to the southern rural area, whereas the northern rural is almost free of disease.
Results
Entomological survey
Results of the entomological survey are shown in Figure 2. In the northern rural area, out of 67 traps, the mean Apparent Density of Tsetse flies (ADT) was 0.62 tsetse/trap/day. In the southern rural area, out of 223 traps mean ADT was substantially higher, at 4.2 tsetse/trap/day. Differences in the number of traps reflect the number of patients in each of the two areas since traps were placed at their main sites of activity. In the northern rural area, out of 51 tsetse flies dissected, one was found infected with Trypanosoma brucei (infection rate 0.197 with Standard Deviation 0.14). In the southern rural area, out of 1496 tsetse flies dissected, 40 were infected (infection rate 0.027 with Standard Deviation 0.15). In the northern rural area, no human blood meal was found out of 5 analyzed blood meals. In the southern rural area, three human blood meals were found out of 87 analyzed blood meals.
Figure 2 Density, infection and human bloodmeals of tsetse flies. This map shows the results of entomological survey. Tsetse flies are far more abundant in the southern rural area than in the northern rural area. Some tsetse flies were caught in the town. All the T. brucei infected tsetse flies were caught in the southern rural area and in the town, except one in the northern rural area. All the bloodmeals taken on humans by tsetse flies were located in the south rural area.
Geographical survey
History of settlement
Before the present town of Bonon was created, the people of the area used to live in 5 distinct villages. The villages of Brozra and Frefredou were located on the present site of the town, near the main road (figure 3) and now constitute districts of Bonon. At the time of colonization, there were not enough people in Brozra and Frefredou to need colonial administration. It was then decided to create a large settlement by moving the inhabitants of the three villages located several kilometers to the north (Séhizra, Vrigrifouta, Zaguié) close to the main road. The resistance of villagers refusing to leave their lands, forced the government to burn the villages in the 1920s, after which the inhabitants settled next to Brozra and Frefredou villages as shown on the topographic map made by the French « Institut de Géographie National » (IGN), from aerial photography taken in 1956 (figure 3). The first non-native ethnic group (Malinke) arrived at the beginning of the twentieth century mainly to trade cola nuts. At that time, the original population was living mainly from hunting and from the cultivation of cash crops (rice and banana). During colonial period, the French government forcibly displaced large numbers of people from Upper Volta (principally the Mossi ethnic group) to work in Côte d'Ivoire, to provide labour for the development of the colony's railways and roads [14]. In 1936, the colonial administration created "villages of colonisation" to house part of the agricultural labour force (Mossi ethnic group) needed to develop the coffee and cocoa plantations in the Central West part of Côte d'Ivoire [15]. With the end of involuntary work (Félix Houphouët-Boigny law, 1946), agricultural laborors were allowed to travel freely. Since 1950, the development of cash crops (cocoa, coffee) in the centre-western region of Côte d'Ivoire (Vavoua, Daloa, Bouaflé, Gagnoa and Koudougou areas), has led to massive immigration of seasonal agricultural labour, mainly originating from Burkina Faso (Mossi, Lobi ethnic groups), Mali (Malinke) and from the north (Senoufo) and the centre (Baoule) of Côte d'Ivoire [16,17]. The native Gouro people gradually gave their lands to the immigrants, usually the parcels situated between their fields and the forest, in order to protect their crops from wild animals (monkeys, elephants). The northern rural area has thus been exploited much earlier than the southern rural area, where there was no established village and only forest.
Figure 3 Location of former villages and Bonon in 1956. The upper map shows the former villages in the north, displaced during the french colonisation in the 1920s. The bottom image shows Bonon in 1956, with the former villages of the North that have been moved near the main road.
Landscape analysis
The departure point of north transect (going towards Bonon) is an encampment surrounded by cocoa and coffee cultivation (figure 4). From there, the landscape opens up, the vegetation becomes lower and less dense and the land becomes flat. Some remnant big trees like the Samba (Triplochiton scleroxylon), the Iroko (Chlorophora excelsa), the Wara (Kola gigantea) and the kapok tree (Ceiba pentandra) testify to the earlier presence of mesophil forest. The transect continues through a succession of fallow burn and fallow with Chromoloena odorata (commonly called « Sékou Touré ») and finishes amongst cashew plantations, banana and manioc cultivation. At the northern periphery of Bonon town, the rice cultivation dominates the low-lying areas. The start of the south transect (going towards Bonon) is in a hilly plantation of coffee mixed with manioc. The line then crosses a small tributary with cocoa and banana trees, which turns into uncultivated lowlands characterized by the presence of bamboo. The principal cultivation is cash and food crops. The vegetation then becomes dominated by very dense herbaceous and shrub layers, which is flowed by a patch of relict forest, with Kola cordifolia, Celtis zincheri and Ceiba pentandra, through which free passage is prevented by a backwater and dense stands of Acacia capensis. Beyond the forest relict, the landscape is dominated by coffee and cocoa plantations. At the end of the transect, approaching the southern periphery of Bonon, the altitude decreases progressively, and the landscape becomes an herbaceous savannah. The major differences between the two transects are therefore the absence of forest and uncultivated low ground along northern transect, and the limited amount of cultivation along the southern line (pictures of figure 4). Analysis of LANDSAT imagery from 2000 shows that this pattern may be extrapolated to the whole study area (figure 4). This shows that, around the town and the villages located along the main road to the east, the original forest environment has been replaced by savannah, Chromoloena odorata fallow and food crops. To the north and west, is the boundary of the Marahoué National Park, which now constitutes a pioneer front for the coffee and cocoa plantations which cover much the landscape, in the north rural area as well as in the south area.
Figure 4 Space structure of study area. On this figure are located the two transects on foot, two pictures of the northern and southern rural areas, and the results of remote sensing analysis. The North and the South transects start in the rural area and go towards the town of Bonon. The pictures of the northern area show that almost all the landscape is exploited by humans. The pictures from the southern area show an association of crops and "natural" vegetation. With the remote sensing analysis results, we can see that all the study area is very exploited by humans and that the relict forest and uncultivated lowgrounds subsist only in the Marahoué national park and in parts of the southern rural area. In the northern rural area, most of the lowgrounds are cultivated. Coffee and cocoa plantations are present everywhere in the rural area, but not close to the biggest human settlements (town and villages).
Human mobility
A total of 81,013 passers-by were counted at the 6 enumeration points (figure 5) over the course of a week, which demonstrates extreme mobility (there are 30,000 people living in the study area, 20,000 in Bonon town and 10,000 in the rural area). Of these, 37,619 were going from the rural area to Bonon and 43,394 from Bonon to the rural area. Out of the 81,013 people counted, 32,739 were in the north rural area (counting points 3, 4 and 5), 23,226 in the south rural area (counting point 1), 14,767 in the west rural area (counting point 2), and 10,281 in the east rural area (counting point 6). Figure 5 shows the mean mobility during the week of counting, for the different tracks and according to the direction taken. In the northern rural area movement is concentrated on three principal tracks. To the south of Bonon, the only track which goes to the south rural area is very heavily used (23,226 people). It is likely people counted one-way were probably counted again on their return; indeed, it was very clear throughout that in the morning, people were going from Bonon to the rural area, and in the evening people were coming back to the town. In figure 6 all observations made are aggregated according to slot time which shows that during the week, the urban population goes to the rural area in the morning and returns to Bonon in the evening, except on Friday (day of market in Bonon) when the rural population goes to town in the morning, and returns to the rural area in the evening. During the week, 1,174 people were questioned at the six counting points. Analysis of the answers shows that some of these people come from or go to other urban centers, mainly Bouaflé town, which seems to have an important link to the Bonon area. Particularly significant is the finding that people going to the northern rural area come mainly from northern neighborhoods, and people going to the southern rural area come mainly from southern neighbourhoods. Movement between rural areas (north, east, west, south) is thus very limited (figure 5) though the urban population is highly mobile.
Figure 5 Mean of daily human mobility between Bonon town and rural space, leaving and destination places of questioned passers-by. This map shows the location of counting points and the results of the human mobility study. The counting points are located around Bonon, on the principal tracks which connect the town and the rural areas. We can see that the mobility in the study area is very important. The south/east track has the highest frequentation and most of the people who take this track live in the south neighborhoods of Bonon or in the southern rural area. We can also see that most of the people who take the northern tracks live in the north neighborhoods of Bonon or in the north rural area.
Figure 6 Mobility between Bonon town and rural space according to direction and hour (mean of the week and Friday). This graph shows the characteristics (according to direction and hour) of mobility between Bonon and the rural area. During the week, people go to work on the morning in the rural area, except Friday (market day), where it is the opposite. The urban mobility is important during the week, but on Friday it is inverted due to the rural people who go to the town of Bonon to sell their products.
Discussion
In this study, we have used information about parasitological, entomological, human and environmental factors, to describe and explain the substantial spatial disparity of HAT between the southern and the northern sectors of the Bonon focus. The results help explain the spatial evolution of HAT prevalence in the Bonon focus since the 1950s. The entomological study undertaken here has shown that in the southern rural area, the density of tsetse flies (Glossina palpalis palpalis) is much higher than in the northern rural area. The places of transmission are apparently located in the southern rural area, mainly in the south-eastern area, frequented by many rural and urban patients, and where numerous tsetse flies (among which 2.67% are infected by T. brucei) were caught. The history of the area's development and its impact on the landscape may be related to the differences in vector density observed between the north and the south rural area. We have shown that at the beginning of the 20th century, the colonial government displaced the inhabitants of the north rural area towards the main road, and also that the south rural was uninhabited. Consequently, the north rural area has been exploited much earlier than the south rural area. The first waves of immigration were into the north area, where immigrants were settled near existing fields. The oldest and biggest settlement in the southern area, the Baoule hamlet called « Deux Côtes », was only created in 1977. In the southern area, contrary to the north, patches of both relict forest and uncultivated lowland remain, being favourable breeding areas for tsetse flies. The risk of trypanosome transmission is not, however, absolutely correlated to vector density, but depends also on a number of contextual factors [18,19] including human mobility. The track which goes to the south-east of Bonon is the most used, and goes through an environment favourable to the human-vector contact (edge of uncultivated lowland/plantations, edge of forest/communication routes) [20]. The transmission of HAT in this area is thus likely to be due to the high density of tsetse flies combined with a high human attendance, in an environment favourable to the human-vector contact. The lack of movement between the north and south is likely to be the main reason to explain that the disease spreads from the south towards the town, but does not spread onward into the densely inhabited northern rural area [8]. This area is thus characterized by a very low presence of vector and parasite, in an environment unfavorable to the human-vector contact, and the chance of transmission is thus too low to enable the development of HAT.
Conclusion
In the focus of Bonon, the history of human settlement has apparently determined to a great extent, both current land cover patterns and the prevailing density of tsetse flies which in turn explain the spatial distribution of HAT. This sequence of events is likely to have happened in other foci within Côte d'Ivoire. Other factors remain, that could not be included in this study but which could also play an important role, such as the presence/absence of reservoir animals, feeding habits of tsetse, the evolution of landuse regarding the socio-political events that have recently occurred in Côte d'Ivoire.
In common with other studies, this work highlights the value of a GIS analysis in the context of multidisciplinary approach, to gaining an understanding of the distribution and spatial dynamics of the disease, and the necessary conditions for effective prevention, notably in Africa [21]. This work, by locating the most active transmission area at the south-east of the focus of Bonon, and by identifying the area with few disease cases, also helps to delineate the boundaries of the focus of Bonon more accurately. Targeted actions of tsetse control starting from north of Bonon (with little chance of tsetse reinvasion due to very low densities) to south, and targeted additional medical surveys in the south will be proposed to the Ivoirian HAT program, in order to reduce sleeping sickness in this area. The methodology used in the present work may be used in other areas of Côte d'Ivoire with similar history and similar environment to predict areas of possible sleeping sickness development/extinction. However it is likely that the current socio-political conditions in Côte d'Ivoire will have an impact on the further spread of HAT, on local, national and even international scales, particularly if forced movements of populations continue to occur [22].
Methods
Study area, presentation and description
The town of Bonon (7°N-6°W) is located about sixty kilometers west of the political capital of the country, Yamoussoukro (figure 1), in the Gouro ethnic group area,. The climate is equatorial with an annual rainfall of around 1 200 mm, and slight annual variations of temperatures (3°C). This region of the Central West Côte d'Ivoire is known for its historical (Bouaflé, Daloa) and contemporary (Vavoua, Sinfra) foci of sleeping sickness, but the focus of Bonon is comparatively recent [13-23]. The town of Bonon was created during the French colonial administration, when populations were forced to settle along communication routes, in order to have direct access to the labour necessary to the development, the maintenance and the exploitation of the area. In the early 70's, Bonon was a big village of agricultural labour immigrants, it has now become a commercial and administrative town (sub-prefecture of Marahoué region) of about 20 000 inhabitants. It is surrounded by villages, hamlets and encampments (with an additional total of about 10 000 inhabitants), representing more than fifty ethnic groups. Due to its infrastructure, Bonon has an important influence in nearby rural areas since the nearest urban centers (Daloa, Bouaflé, Sinfra) are located more than 30 kilometers away.
Entomological survey
A total of 290 Vavoua traps [24] were set up on each patient's place of residence, water supply sites and working places- with the aim of obtaining data on the vectors distributions in relation human presence. Apparent Density per Trap per day (ADT: number of tsetse flies caught per trap and per day), infection by Trypanosoma brucei (the pathogenic parasite) using molecular tools (see below), and number of bloodmeals taken from humans were monitored. In the northern part (almost free of disease), traps were set up in areas the patients frequented. Each trap remained in place for four days, with cages being changed daily, and fly counts, sex-ratio determinations, and dissections being carried out daily. Tsetse flies (G. palpalis) were dissected to look for trypanosomes in the mouthparts, midgut, and salivary glands. Each organ (mouthparts, salivary glands, midgut) was put into a separate eppendorf tube containing 30 μl sterile distilled water for subsequent molecular analyses. Trypanosoma brucei identification was done by molecular analysis (PCR, polymerase Chain Reaction) using specific primers [25]. In order to assess levels of contact between human and tsetse, the bloodmeal origin was assessed using the method of Diallo et al. [26]. This method, based on variation of the Super Oxyde Dismutase (SOD) enzyme, distinguishes between bloodmeals taken from humans and animals.
Geographical survey
Historical settlement of study area
Five administrative workers and several community leaders (religion, ethnic groups, youth) of town of Bonon and villages were interviewed, alternatively using individual or collective (around 20 people) discussions, in order to understand the history of the progressive settlement in this area. These interviews took place after several preliminary meetings in which the aim of the work was explained, in order to avoid any confusion due to the special context which prevailed in Côte d'Ivoire at the time of the study (March to May 2004). An additional objective was also to understand the evolution of the relationships between the different communities.
Landscape analysis
Two transects were carried out on foot (figure 4), one in the northern rural area of Bonon, the other in the southern rural area. The lines of these two transects were defined according to several parameters:
-Analysis of satellite image (Landsat 2000) covering the study area, which was used to determined landscape discontinuity.
-Knowledge of the area, acquired during the multiple travels by car and motorcycle.
-Where the patients lived.
These two transects were each 7 kilometers long. Landscape characteristics were assessed at points 250 metres apart using a description form, on which Global Positioning System (GPS) coordinates, landscape details, principal cultivation types and the names of the principal plant species were recorded. The data obtained were used to draw up a thematic classification of landscape, which was then extrapolated to the whole study area by the signal processing of Landsat 2000 satellite image. This signal processing was done under ENVI 3.2 (Environment for Visualizing images) software.
Human mobility
We defined six counting points (figure 5) on the principal tracks which link Bonon to the surrounding rural areas. At each of these points, every individual going from Bonon to the rural area and from the rural area to Bonon, from 6 am to 7 pm, every day during one week was recorded. At the same time, a random subsample was asked to complete a questionnaire, in order to obtain more detailed information on the nature of their journeys (place of departure, place of destination, reason for travelling).
All the data were recorded in a linked data base under Access. The data were then imported into Arcview 3.2., which was used to execute the spatial queries.
Authors' contributions
Conception of the study: PS, JPD, VJ, FC.
Data acquisition, analysis, interpretation: FC, EO, BC, YO, VJ, SD, PS.
Critical revision of the manuscript: PS, VJ, GC, JPD.
All authors read and approved the final manuscript.
Acknowledgements
We are extremely grateful to the Ivoirian HAT control program and to the team « THA et glossines » of Institut Pierre Richet for their help, and to the national and regional administrative authorities for their collaboration. This work was funded by IRD and SCAC (Service de Coopération et d'Action Culturelle) Abidjan. We thank very much Dr. W. WINT (Environmental Research Group Oxford, UK) for critical reading of the manuscript.
==== Refs
World Health Organization African trypanosomosis WHO Report on Global Surveillance of Epidemic-prone Infectious Diseases 2001
Cattand P L'épidémiologie de la trypanosomiase humaine africaine: une histoire multifactorielle complexe Med Trop 2001 61 313 317
Picheral H CEGET (CNRS) Complexes et systèmes pathogènes: approche géographique de l'épidémiologie à la géographie humaine, travaux et documents de géographie tropicale 1983 48 Talence: Doumenge 5 22
Hervouët JP Laveissière C Les grandes endémies: l'espace social coupable Polit Afric 1987 28 21 32
Laveissière C Hervouët JP La trypanosomiase humaine en Afrique de l'Ouest Epidémiologie et contrôle 1991 ORSTOM, Collection Didactique
Reid R Kruska RL Deitchmann U Thorthon PK Leak SGA Human population growth and the extinction of the tsetse fly Agric Ecosyst Env 2000 77 227 236 10.1016/S0167-8809(99)00103-6
Gouteux JP Nkouka E Noireau F Frézil JL Sinda D The tsetse-flies of BrazzavilleI. Repartition and importance of the breeding and resting sites Rev Elev Méd Vét pays Trop 1986 39 355 362
Courtin F Dupont S Zeze DG Jamonneau V Sane B Coulibaly B Cuny G Solano P Trypanosomose Humaine Africaine: transmission urbaine dans le foyer de Bonon (Côte d'Ivoire) Trop Med Int Hlth 2005 4 340 346 10.1111/j.1365-3156.2005.01398.x
Hervouët JP Laffly D Cardon L La maladie du sommeil en Côte d'Ivoire: à la recherche d'indicateurs de risque Espace Popul Soc 2000 2 209 225
Hervouët JP Laveissière C CEGET (CNRS) Les interrelations homme/milieu/glossines et leurs répercussions sur le développement de la maladie du sommeil en Côte d'Ivoire de l'épidémiologie à la géographie humaine, travaux et documents de géographie tropicale 1983 48 Talence: Doumenge 139 147
Lointier M Truc P Drapeau L Nanga S Tarek M Méthodologie de détermination de zones à risque de maladie du sommeil en Côte d'Ivoire par approche spatialisée Méd Trop 2001 61 390 396
Laffly D Hervouët JP Une mouche tsé-tsé dans le capteur ! Identification de facteurs de risque de la Trypanosomiase Humaine Africaine par télédétection et analyse spatiale Espace Popul Soc 2000 2 227 240
Solano P Kone A Garcia A Sane B Michel V Michel JF Coulibaly B Jamonneau V Kaba D Dupont S Fournet F Rôle des déplacements des malades dans l'épidémiologie de la Trypanosomose Humaine Africaine dans le foyer de Bonon, Côte d'Ivoire Méd Trop 2003 63 577 582
Londres A terre d'ébène, la traite des noirs 1929 Paris, A. Michel
Mandé I Les migrations du travail en Haute-Volta (actuel Burkina Faso), mise en perspective historique (1919–1960) PhD thesis 1997 Université Denis Diderot Paris 7, U.F.R. Géographie Histoire et sciences de la Société
Chauveau JP L'économie de plantation villageoise caféière et cacaoyère en Côte d'Ivoire:les leçons de l'histoire 1985 8 Les cahiers de la Recherche-Développement 46 49
Balac R Karthala Dynamiques migratoires et économie de plantations La Côte d'Ivoire à l'aube du XXI siècle, Défis démographiques et développement durable Paris 2002 195 231
Nash TAM Tsetse flies of British West Africa 1948 O.B.E., D. Sc. (London)
Laveissière C Grebaut P Herder S Penchenier L Les glossines vectrices de la trypanosomiase humaine africaine 2000 IRD/OCEAC, Yaoundé
Laveissière C Couret D Hervouët JP Localisation et fréquence du contact homme/glossine en secteur forestier de Côte d'ivoire, Recherche des points épidémiologiquement dangereux dans l'environnement végétal Cah ORSTOM, Sér Ent méd Parasitol 1986 24 21 35
Tanser FC le Sueur D The application of geographical information systems to important public health problems in Africa Int J Health Geogr 2002 l 1 9
Kaba D Dje NN Courtin F Oké E Koffi M Garcia A Jamonneau V Solano P Impact des évènements socio-politiques sur l'évolution de la THA dans le centre-ouest de la Côte d'Ivoire Trop Med Int Hlth
Djè NN Miezan TW N'Guessan P Brika P Doua F Boa F Distribution géographique des trypanosomés pris en charge en Côte d'Ivoire de 1993 à 2000 Bull soc Path Exo 2002 95 359 361
Laveissière C Grébaut P Recherches sur les pièges à glossines: mise au point d'un modèle économique: le piège "Vavoua" Trop Med Parasitol 1990 41 185 192 2166330
Moser DR Cook GA Ochs DE Bailey CP McKane MR Donelson JE Detection of Trypanosoma congolense and Trypanosoma brucei subspecies by DNA amplification using the polymerase chain reaction Parasitology 1989 99 57 66 2797872
Diallo BP Truc P Laveissière C A new method for identifying blood meals of human origin in tsetse flies Acta Trop 1997 63 61 64 9083586 10.1016/S0001-706X(97)86626-0
|
16269078
|
PMC1308845
|
CC BY
|
2021-01-04 16:39:05
|
no
|
Int J Health Geogr. 2005 Nov 3; 4:27
|
utf-8
|
Int J Health Geogr
| 2,005 |
10.1186/1476-072X-4-27
|
oa_comm
|
==== Front
Immun AgeingImmunity & ageing : I & A1742-4933BioMed Central London 1742-4933-2-141627114310.1186/1742-4933-2-14Short ReportOsteoporosis, inflammation and ageing Ginaldi Lia [email protected] Benedetto Maria Cristina [email protected] Martinis Massimo [email protected] Department of Internal Medicine, University of L'Aquila, Italy2005 4 11 2005 2 14 14 31 8 2005 4 11 2005 Copyright © 2005 Ginaldi et al; licensee BioMed Central Ltd.2005Ginaldi et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Osteoporosis is a condition characterized by low bone mass and increased bone fragility, putting patients at risk of fractures, which are major causes of morbidity substantially in older people. Osteoporosis is currently attributed to various endocrine, metabolic and mechanical factors. However, emerging clinical and molecular evidence suggests that inflammation also exerts significant influence on bone turnover, inducing osteoporosis. Numerous proinflammatory cytokines have been implicated in the regulation of osteoblasts and osteoclasts, and a shift towards an activated immune profile has been hypothesized as important risk factor. Chronic inflammation and the immune system remodelling characteristic of ageing, as well as of other pathological conditions commonly associated with osteoporosis, may be determinant pathogenetic factors. The present article will review the current perspectives on the interaction between bone and immune system in the elderly, providing an interpretation of osteoporosis in the light of inflamm-ageing.
AgeingInflammationCytokinesBone remodellingOsteoporosis
==== Body
Introduction
Osteoporosis is a relevant age related disorder, characterized by low bone mass and increased bone fragility putting the patient at risk for fractures [1]. Currently osteoporosis is viewed as a heterogeneous condition which can occur in any age of life and its etiology is attributed to various endocrine, metabolic and mechanical factors (abnormalities of parathyroid hormone and calcitonin secretion, insufficient vitamin D and calcium intake, postmenopausal hormonal condition, pregnancy, nutritional disorders, immobility and consumption of drugs such as cortisone, among others) [2]. Recently, growing understanding of the bone remodelling process suggests that factors involved in inflammation are linked with those critical for bone physiology and remodelling, supporting the theory that inflammation significantly contributes to the aetiopathogenesis of osteoporosis [3,4].
Is osteoporosis an inflammatory process?
Clinical observations reveal coincidence of systemic osteoporosis with period of systemic inflammation as well as co-localization of regional osteoporosis with areas of regional inflammation [2]. Different epidemiologic studies report an increase in the risk of developing osteoporosis in various inflammatory conditions [5-8]. Immunological dysfunctions, autoimmune and chronic inflammatory diseases [9], HIV infection [10], hyper-IgE syndrome [11], rheumatoid arthritis [12], haematological diseases, particularly myeloma [13], and inflammatory bowel diseases [14], are associated with osteoporosis.
Erosions seen in conditions such as gout, osteomyelitis, rheumatoid arthritis, ankylosing spondylitis, and psoriatic arthritis, are typically associated with inflammation in the joints. Pro-osteoclastic cytokines, such as tumour necrosis factor (TNF)-α and interleukin (IL)-6, are elevated in these conditions and local cytokine profile is consistent with the cytokines that modulate bone resorption [15,16].
C-reactive protein (CRP) production in the liver is upregulated by IL-1, IL-6 and TNF-α, and is regarded as a sensitive marker of systemic inflammation [17,18]. An association between circulating high sensitive (hs)CRP level and bone mineral density has been observed in several immune and inflammatory diseases, as well as in healthy individuals, suggesting a relationship between subclinical systemic inflammation and osteoporosis [19].
Rheumatoid arthritis (RA) is a typical example of the link between inflammation and osteoporosis. Bone loss in RA occurs both in the joints and throughout the skeleton as a result of the release of proteinases (metalloproteinases) and proinflammatory cytokines (IL-1, TNF-α), which are responsible for cartilage and bone destruction. As a result, disease activity is an independent risk factor for osteoporosis in RA [20].
A temporal link between inflammation and osteoporosis also emerges in conditions such as ageing, menopause, pregnancy, transplantation and steroid administration. While nutritional, mechanical and hormonal factors clearly play a role in many of these situations, the concordance of osteoporosis and inflammation is buttressed by emerging molecular evidence of mediating immunological factors [2].
On the other hand, one intriguing aspect of immunosenescence is the increased production of pro-inflammatory cytokines with age and a close link between age-related systemic inflammatory process (inflamm-ageing)[21] and osteoporosis is well documented [22].
A molecular scenario of immune regulated bone loss
Although osteoporosis is not typically considered an immunological disorder, recent data have indicated overlapping pathways between bone biology and biology of inflammation [23-25]. Certain pro-inflammatory cytokines play potential critical roles both in the normal bone remodelling process and in the pathogenesis of perimenopausal and late-life osteoporosis [26]. For example, interleukin (IL)-6 promotes osteoclast differentiation and activation [27]. This cytokine is involved in the pathogenesis of various metabolic bone diseases, including postmenopausal osteoporosis, Paget's disease and osteoporosis associated with heamtologic malignancies [28]. IL-1 is another potent stimulator of bone resorption [29] that has been linked to the accelerated bone loss seen in idiopathic and postmenopausal osteoporosis [30]. TNF-α is implicated in tumour-induced bone resorption and non-tumour-induced osteopenia [31]. Anti-TNF drugs, currently used in the therapy of several immunological disorders, are also useful in preventing and/or reversing systemic bone loss associated to the disease, as they target both the bone and the inflammatory process [20]. The production of IL-1, IL-6, and/or TNF-α by peripheral blood monocytes has been positively correlated with bone resorption or spinal bone loss in healthy pre- and postmenopausal women [32].
The inflammatory mediator nitric oxide (NO) is also involved in the pathogenesis of osteoporosis. The activation of the inducible NO synthesis (iNOS) pathway by cytokines, such as IL-1 and TNF-α, inhibits osteoblast function in vitro and stimulates osteoblast apoptosis [33].
The TNF-family molecule RANKL and its receptor RANK (receptor activator of NFkB ligand) have been specifically implicated in the bone loss in rheumatoid arthritis [34]. They are key regulators of bone remodelling and are essential for the development and activation of osteoclasts. Intriguingly, RANKL/RANK interactions also regulate T cell/dendritic cell communication, dendritic cell survival and lymphnode formation [35]. Calciotropic factors such as vitamin D3, PGE2, IL-1, IL-11, TNF-α and glucocorticoid induce RANKL expression on osteoblasts [36]. RANKL binding to the RANK expressed on haematopoietic progenitors activates a signal transduction cascade that leads to osteoclast differentiation. Moreover, RANKL stimulates bone resorbing activity in mature osteoclasts via RANK. When RANK is activated, it sends signals into the cells through tumor necrosis factor receptor-associated factors (TRAFs), mainly TRAF6. These RANK associated molecules, through downstream pathways such as NF-kB, JNK/SAPK, p38 and Akt/PKB, regulate bone resorption, activation, survival, and differentiation of osteoclasts and dendritic cells.
Osteoprotegerin (OPG), also known as osteoclastogenesis inhibitory factor, functions as a soluble decoy receptor to RANKL and competes with RANK for RANKL binding. TGFβ released from bone during active bone resorption has been suggested as a feedback mechanism by upregulating OPG level. Estrogen can enhance OPG production on osteoblasts which is a possible explanation of postmenopausal osteoporosis following estrogen withdrawal [35]. Since the expression of RANKL/RANK can be controlled by sex hormones, it is possible to speculate that this system may control gender specific differences in immunity and could be involved in the higher incidence of autoimmune diseases and osteoporosis in women. RANK-L, RANK and OPG therefore configure interesting molecular links between bone remodelling, immunity and inflammation.
The emergence of osteoimmunology
There exists an intimate interplay between the bone and the immune system which has been termed osteoimmunology [3]. The activity of immune cells affects the balance of bone mineralization and resorption carried out by the opposing actions of osteoblasts and osteoclasts [37].
Dendritic cells, specialized to present antigens, and osteoclasts, specialized to resorb bone, share the same bone marrow precursors of the monocyte lineage and exhibit parallel lifecycles, regulated by a variety of cytokines, transcription factors and inflammatory mediators. Molecules that regulate osteoclastogenesis are key factors in many immunological functions. For example, TRAF6 functions as a molecular bridge spanning adaptive immunity, innate immunity and osteoimmunology [38]. TRAF6-deficient mice lack osteoclasts and concomitantly have defects in cytokine production and T cell stimulation [39].
Activated T cells affect bone physiology producing cytokines that lead to RANKL expression on osteoblasts. Moreover, activated T cells directly express and produce RANKL that induces osteoclast formation and activation through its specific receptor [28,35]. There are multiple mechanisms and interactions by which cytokines regulate bone resorption. IL-6 contributes to RANKL upregulation in osteoblastic cells. IL-1 and TNF-α may not only promote osteoclast generation, but they also appear to stimulate mature osteoclasts to perform more resorption cycles through modulation of RANKL activity. IL-1 is further involved in bone metabolism as an osteoblast activator: osteoblasts secrete RANKL which promotes survival and differentiation of the osteoclast precursors to mature osteoclasts through RANK. IL-1 and IL-6 also directly enhance osteoclast activity by RANKL-independent mechanisms. They may directly extend the lifespan of the osteoclasts by inhibiting osteoclast apoptosis. Both TNF-α and IL-1 inhibit collagen synthesis in osteoblasts and enhance degradation of the extracellular matrix [18]. In inflammatory or autoimmune disease states, activated T cells produce RANKL and pro-inflammatory cytokines, all of which can induce RANKL expression in osteoblasts [40]. However, the constant activity of T cells fighting the universe of antigens to which we are exposed does not usually cause extensive bone loss. Multiple T cell-derived cytokines, such as IL-12 and IL-18, might be able to interfere with RANK signalling and therefore with osteoclastogenesis and osteoclast functions. A crucial counter-regulatory mechanism whereby activated T cells can inhibit osteoclast development and activation is through the action of the antiviral cytokine interferon (IFN)-γ. Mechanistically, IFN-γ activates the ubiquitin-proteasome pathway within the osteoclasts, resulting in the degradation of TRAF6 [38].
Ageing and osteoporosis: immunological links
Inflamm-ageing may, at least partially, be a common mechanism for the development of lower bone mass and other age-related disorders [41,42]. Senility is in fact notable for acceleration of diseases that are increasingly attributed to inflammation, such as atherosclerosis, Alzheimer's disease and asthma [21,43,44]. Many cytokines, including IL-6, TNF-α, IL-1, are elevated during senescence and play direct roles in the pathogenesis of these diseases [45]. All these cytokines are stimulators of osteoclast activity as well. Associations between atherosclerosis and osteopenia have been well documented [46,47]. These findings suggest a potential causal relationship between systemic inflammation that is observed in the elderly and the prevalence of generalized age-related osteoporosis [2]. Moreover, the increased catabolic signal, driven by inflammation also in the absence of clinically diagnosable inflammatory diseases [44], could be able to induce osteoblast apoptosis [48], as well as apoptosis of muscle cells, leading to age-related osteoporosis and sarcopenia respectively.
During ageing, under the influence of the lifelong exposure to chronic antigenic load and oxidative stress, the physiological counter-regulatory process which inhibits bone resorption following T cell activation is likely impaired. This would contribute, together with the age-related systemic low-grade inflammation, to the increasing incidence of osteoporosis during senescence.
Osteoporosis, as well as other age-related disorders, has a strong genetic component and the rate extent of bone loss during senescence varies widely between individuals. It is tempting to postulate that this may be in part due to individual differences in cytokine activity. In support of this hypothesis it has been demonstrated that certain IL-6 polymorphisms are able to influence the risk of osteoporosis in postmenopausal women [49]. Similarly, IL-1β and IL-1 receptor antagonist (IL-1Ra) gene polymorphisms are associated with reduced bone mineral density and predispose women to osteoporosis at the lumbar spine [50].
Moreover, also the sex hormonal decline which accompanies ageing contributes to the pathogenesis of senile osteoporosis through immunologically mediated mechanisms. It has been postulated that estrogens exert their effect on bone not only by direct action per se, but also by inhibiting IL-6 gene expression. A similar relationship between androgen and IL-6 gene expression also exists [51]. The decline in ovarian function is associated with decreased OPG production and spontaneous increases in proinflammatory and pro-osteoclastic cytokines such as IL-6, TNF-α, and IL-1 [32,52,53].
Osteoporosis within an evolutionary perspective
During evolution of multi-cellular organisms, the increasing complexity of emerging systemic functions, such as inflammation, may have led to a widening range of demand for key minerals such as calcium and phosphate and storage functions may have evolved to provide mineral reservoirs [2]. Moreover, calcium is an important component of milk and its transport from mother to the fetus and neonate is a vital process to preserve species. Intriguingly, RANKL and RANK also play essential roles in the formation of a lactating mammary gland in pregnancy. This could partly contribute to both the immune remodelling and the accelerated bone loss during pregnancy and lactating [35]. Bone can thus be viewed as a mobilizable reservoir of calcium and phosphate as salts. Over time, the regulation of bone turnover was probably optimized evolutionarily to address the combined metabolic and structural demands of the host.
In this perspective osteoporosis may reflect a state of disequilibrium between structural demand for calcium and phosphate and their biologic demand during metabolically active states such as inflammation. Inflammation is the leading force driving immunosenescence and the chronic low-grade inflammatory state characteristic of ageing represents the predisposing substrate on which osteoporosis, as well as other age-related diseases, might emerge [21,54].
The lack of evolutionary selective pressure post-reproduction may have further exacerbated this disequilibrium in modern times and many biologic functions acquired during evolution have become maladaptive. Presently, in most developed countries, the human lifespan is greatly increased, and many individuals are living into post-reproductive senescence, an evolutionarily naive life epoch. Obesity, diabetes, Alzheimer's disease and atherosclerosis are examples of diseases of modernity that are attributable to modern living circumstances or that are unmasked during senility, and also the emergence of osteoporosis as a modern disease may be an example of this phenomenon [2,21,44].
==== Refs
Delmas PD Treatment of postmenopausal osteoporosis Lancet 2002 359 2018 26 12076571 10.1016/S0140-6736(02)08827-X
Yun AJ Lee PY Maladaptation of the link between inflammation and bone turnover may be a key determinant of osteoporosis Med Hypotheses 2004 63 532 537 15288383 10.1016/S0306-9877(03)00326-8
Arron JR Choi Y Bone versus immune system Nature 2000 408 535 6 11117729 10.1038/35046196
Lorenzo J Interactions between immune and bone cells: new insights with many remaining questions J Clin Invest 2000 106 749 52 10995785
Mitra D Elvins DM Speden DJ Collins AJ The prevalence of vertebral fractures in mild ankylosing spondylitis and their relationship to bone mineral density Rheumatology 2000 39 85 9 10662879 10.1093/rheumatology/39.1.85
Haugeberg G Lodder MC Lems WF Uhlig T Orstavik RE Dijkmans BA Kvien TK Woolf AD Hand cortical bone mass and its associations with radiographic joint damage and fractures in 50–70 year old female patients with rheumatoid arthritis: cross sectional Oslo-Truro-Amsterdam (OSTRA) collaborative study Ann Rheum Dis 2004 63 1331 4 15361395 10.1136/ard.2003.015065
Bultink IE Lems WF Kostense PJ Dijkmans BA Voskuyl AE Prevalence of and risk factors for low bone mineral density and vertebral fractures in patients with systemic lupus erythematosus Arthritis Rheum 2005 52 2044 50 15986345 10.1002/art.21110
Mikuls TR Saag KG Curtis J Bridges SL JrAlarcon GS Westfall AO Lim SS Smith EA Jonas BL Moreland LW CLEAR investigators Prevalence of osteoporosis and osteopenia among African Americans with early rheumatoid arthritis: the impact of ethnic-specific normative data J Natl Med Assoc 2005 97 1155 60 16173331
Walsh NC Gravallese EM Bone loss in inflammatory arthritis: mechanisms and treatment strategies Curr Opin Rheumatol 2004 16 419 27 15201606 10.1097/01.bor.0000127824.42507.68
Annapoorna N Rao GV Reddy NS Rambabu P Rao KR An increased risk of osteoporosis during acquired immunodeficiency syndrome Int J Med Sci 2004 1 152 164 15912194
Leung DY Geha RS Clinical and immunologic aspects of the hyperimmunoglobulin E syndrome Hematol Oncol Clin North Am 1988 2 81 100 3279022
Jensen T Klarlund M Hansen M Jensen KE Skjodt H Hydlldstrup L Connective tissue metabolism in patients with unclassified polyarthritis and early rheumatoid arthritis. Relationship to disease activity, bone mineral density, and radiographyc outcome J Rheumatol 2004 31 1698 1708 15338487
Abrahamsen B Andersen I Christensen SS Skov Madsen J Brixen K Utility of testing for monoclonal bands in serum of patients with suspected osteoporosis: retrospective, cross sectional study BMJ 2005 330 818 15767267 10.1136/bmj.38376.401701.8F
Moschen AR Kaser A Enrich B Ludwiczek O Gabriel M Obrist P Wolf AM Tilg H The RANKL/OPG system is activated in inflammatory bowel disease and relates to the state of bone loss Gut 2005 54 479 87 15753532 10.1136/gut.2004.044370
Ishihara K Hirano T IL-6 in autoimmune disease and chronic inflammatory proliferative disease Cytokine Growth Factor Rev 2002 13 357 68 12220549 10.1016/S1359-6101(02)00027-8
Moschen AR Kaser A Enrich B Ludwiczek O Gabriel M Obrist P Wolf AM Tilg H RANKL/OPG system is activated in inflammatory bowel disease and relates to the state of bone loss Gut 2005 54 479 87 15753532 10.1136/gut.2004.044370
Koh JM Kang YH Jung CH Bae S Kim DJ Chung YE Kim GS Higher circulating hsCRP levels are associated with lower bone mineral density in healthy pre-and postmenopausal women: evidence for a link between systemic inflammation and osteoporosis Osteoporosis Int 2005 doi 10.1007/s00198-005-1840-5
Muller B Cytokine imbalance in non immunological chronic disease Cytokine 2002 18 334 339 12160522 10.1006/cyto.2002.0882
Ganesan K Teklehaimanot S Tran TH Asuncion M Norris K Relationship of C-reactive protein and bone mineral density in community-dwelling elderly females J Natl Med Assoc 2005 97 329 33 15779496
Saidenberg-Kermanac'h N Cohen-Solal M Bessis N De Vernejoul MC Boissier MC Role for osteoprotegerin in rheumatoid inflammation Joint Bone Spine 2004 71 9 31 14769514 10.1016/S1297-319X(03)00131-3
Franceschi C Bonafè M Valensin S Olivieri F De Luca M Ottaviani E De Benedictis G Inflamm-ageing. An evolutionary perspective on immunosenescence Ann N Y Acad Sci 2000 908 244 54 10911963
Kiecolt-Glaser JK Preacher KJ MacCallum RC Atkinson C Malarkey WB Glaser R Chronic stress and age-related increases in the proinflammatory cytokine IL-6 Proc Natl Acad Sci USA 2003 100 9090 5 12840146 10.1073/pnas.1531903100
Goldring SR Inflammatory mediators as essential elements in bone remodeling Calcif Tissue Int 2003 73 97 100 14565589 10.1007/s00223-002-1049-y
Siggelkow H Eidner T Lehmann G Viereck V Raddarz D Munzel U Hein G Hufner M Cytokines, osteoprotegerin, and RANKL in vitro and histomorphometric indices of bone turnover in patients with different bone diseases J Bone Miner Res 2003 18 529 38 12619938
Pfeilschifter J Role of cytokines in postmenopausal bone loss Curr Osteoporos Rep 2003 1 53 8 16036065
Manolagas SC Jilka RL Bone marrow, cytokines, and bone remodeling. Emerging insights into the pathophysiology of osteoporosis N Engl J Med 1995 332 305 11 7816067 10.1056/NEJM199502023320506
Manolagas SC Birth and death of bone cells: basic regulatory mechanisms and implications for the pathogenesis and treatment of osteoporosis Endocr Rev 2000 21 115 137 10782361 10.1210/er.21.2.115
Takayanagi H Ogasawara K Hida S Chiba T Murata S Sato K Takaoka A Yokochi T Oda H Tanaka K Nakamura K Taniguchi T T-cell-mediated regulation of osteoclastogenesis by signalling cross-talk between RANKL and IFN-gamma Nature 2000 408 600 5 11117749 10.1038/35046102
Wei S Kitaura H Zhou P Ross FP Teitelbaum SL IL-1 mediates TNF-induced osteoclastogenesis J Clin Invest 2005 115 282 90 15668736 10.1172/JCI200523394
Pacifici R Rifas L McCracken R Vered I McMurtry C Avioli LV Peck WA Ovarian steroid treatment blocks a postmenopausal increase in blood monocyte interleukin 1 release Proc Natl Acad Sci USA 1989 86 2398 2402 2522659
Moffett SP Zmuda JM Oakley JI Beck TJ Cauley JA Stone KL Lui LY Ensrud KE Hillier TA Hochberg MC Morin P Peltz G Greene D Cummings SR Tumor necrosis factor-alpha polymorphism, bone strength phenotypes, and the risk of fracture in older women J Clin Endocrinol Metab 2005 90 3491 7 15797957 10.1210/jc.2004-2235
Scheidt-Nave C Bismar H Leidig-Bruckner G Woitge H Seibel MJ Ziegler R Pfeilschifter J Serum interleukin 6 is a major predictor of bone loss in women specific to the first decade post menopause J Clin Endocrinol Metab 2001 86 2032 42 11344203 10.1210/jc.86.5.2032
Armour KJ Armour KE van't Hof RJ Reid DM Wei XQ Liew FY Ralston SH Activation of the inducible nitric oxide synthase pathway contributes to inflammation-induced osteoporosis by suppressing bone formation and causing osteoblast apoptosis Arthritis Rheum 2001 44 2790 6 11762939 10.1002/1529-0131(200112)44:12<2790::AID-ART466>3.0.CO;2-X
Romas E Gillespie MT Martin TJ Involvement of receptor activator of NFkappaB ligand and tumor necrosis factor-alpha in bone destruction in rheumatoid arthritis Bone 2002 30 340 6 11856640 10.1016/S8756-3282(01)00682-2
Theill LE Boyle WJ Penninger JM KANK-L and RANK: T cells, bone loss, and mammalian evolution Annu Rev Immunol 2002 20 795 823 11861618 10.1146/annurev.immunol.20.100301.064753
Eghbali-Fatourechi G Khosla S Sanyal A Boyle WJ Lacey DL Riggs BL Role of RANK ligand in mediating increased bone resorption in early postmenopausal women J Clin Invest 2003 111 1221 30 12697741 10.1172/JCI200317215
Teitelbaum SL Osteoclasts, integrins, and osteoporosis J Bone Miner Metab 2000 18 344 9 11052468 10.1007/s007740070007
Wu H Arron JR TRAF6, a molecular bridge spanning adaptive immunity, innate immunity and osteoimmunology BioEssays 2003 25 1096 1105 14579250 10.1002/bies.10352
Lomaga MA Yeh WC Sarosi I Duncan GS Furlonger C Ho A Morony S Capparelli C Van G Kaufman S van der Heiden A Itie A Wakeham A Khoo W Sasaki T Cao Z Penninger JM Paige CJ Lacey DL Dunstan CR Boyle WJ Goeddel DV Mak TW TRAF6 deficiency results in osteopetrosis and defective interleukin-1, CD40, and LPS signaling Genes Dev 1999 13 1015 24 10215628
Hofbauer LC Lacey DL Dunstan CR Spelsberg TC Riggs BL Khosla S Interleukin-1beta and tumor necrosis factor-alpha, but not interleukin-6, stimulate osteoprotegerin ligand gene expression in human osteoblastic cells Bone 1999 25 255 9 10495128 10.1016/S8756-3282(99)00162-3
Wallin R Wajih N Greenwood GT Sane DC Arterial calcification: a review of mechanisms, animal models, and the prospects for therapy Med Res Rev 2001 21 274 301 11410932 10.1002/med.1010
Hirose K Tomiyama H Okazaki R Arai T Koji Y Zaydun G Hori S Yamashina A Increased pulse wavevelocity associated with reduced calcaneal quantitative osteo-sono index: possible relationship between atherosclerosis and osteopenia J Clin Endocrinol Metab 2003 88 2573 2578 12788857 10.1210/jc.2002-021511
Bruunsgaard H Pedersen BK Age-related inflammatory cytokines and disease Immunol Allergy Clin North Am 2003 23 15 39 12645876 10.1016/S0889-8561(02)00056-5
Ginaldi L De Martinis M Monti D Franceschi C Chronic antigenic load and apoptosis in immunosenescence Trends Immunol 2005 26 79 84 15668122 10.1016/j.it.2004.11.005
Bruunsgaard H Effects of tumor necrosis factor-alpha and interleukin-6 in elderly populations Eur Cytokine Netw 2002 13 389 391 12517724
Hak AE Pols HA van Hemert AM Hofman A Witteman JC Progression of aortic calcification is associated with metacarpal bone loss during menopause: a population-based longitudinal study Arterioscler Thromb Vasc Biol 2000 20 1926 1931 10938013
Jorgensen L Engstad T Jacobsen BK Bone mineral density in acute stroke patients: low bone mineral density may predict first stroke in women Stroke 2001 32 47 51 11136913
Roubenoff R Catabolism of aging: is it an inflammatory process? Curr Opin Clin Nutr Metab Care 2003 6 295 299 12690262 10.1097/00075197-200305000-00005
Ferrari SL Ahn-Luong L Garnero P Humphries SE Greenspan SL Two promoter polymorphisms regulating interleukin-6 gene expression are associated with circulating levels of C-reactive protein and markers of bone resorption in postmenopausal women J Clin Endocrinol Metab 2003 88 255 259 12519862 10.1210/jc.2002-020092
Chen HY Chen WC Wu MC Tsai FJ Lin CC Interleukin-1beta and interleukin-1 receptor antagonist gene polymorphism in postmenopausal women: correlation to bone mineral density and susceptibility to osteoporosis Maturitas 2002 44 49 54 12568735 10.1016/S0378-5122(02)00313-4
Liu H Liu K Bodenner DL Estrogen receptor inhibits interleukin-6 gene expression by disruption of nuclear factor kappa B transactivation Cytokine 2005 31 251 257 16043358 10.1016/j.cyto.2004.12.008
Pfeilschifter J Koditz R Pfohl M Schatz H Changes in proinflammatory cytokine activity after menopause Endocr Rev 2002 23 90 119 11844745 10.1210/er.23.1.90
Ershler W Harman SM Keller ET Immunologic aspects of osteoporosis Dev Comp Immunol 1997 21 487 499 9463782 10.1016/S0145-305X(97)00029-3
De Martinis M Franceschi C Monti D Ginaldi L Inflamm-ageing and lifelong antigenic load as major determinants of ageing rate and longevity FEBS Lett 2005 579 2035 2039 15811314 10.1016/j.febslet.2005.02.055
|
16271143
|
PMC1308846
|
CC BY
|
2021-01-04 16:36:32
|
no
|
Immun Ageing. 2005 Nov 4; 2:14
|
utf-8
|
Immun Ageing
| 2,005 |
10.1186/1742-4933-2-14
|
oa_comm
|
==== Front
Int Semin Surg OncolInternational seminars in surgical oncology : ISSO1477-7800BioMed Central London 1477-7800-2-251627765510.1186/1477-7800-2-25ReviewLymphoscintigraphy and triangulated body marking for morbidity reduction during sentinel node biopsy in breast cancer Krynyckyi Borys R [email protected] Michail K [email protected] Suk Chul [email protected] Dong Wook [email protected] Arlene [email protected] Renee M [email protected] Chun K [email protected] Department of Radiology, Division of Nuclear Medicine, The Mount Sinai School of Medicine, The Mount Sinai Hospital, New York, New York, USA2 Department of Surgery, The Mount Sinai School of Medicine, The Mount Sinai Hospital, New York, New York, USA3 Department of Nuclear Medicine, Albert Einstein College of Medicine of Yeshiva University, and the Montefiore Medical Center, Bronx, New York, USA2005 8 11 2005 2 25 25 12 10 2005 8 11 2005 Copyright © 2005 Krynyckyi et al; licensee BioMed Central Ltd.2005Krynyckyi et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current trends in patient care include the desire for minimizing invasiveness of procedures and interventions. This aim is reflected in the increasing utilization of sentinel lymph node biopsy, which results in a lower level of morbidity in breast cancer staging, in comparison to extensive conventional axillary dissection. Optimized lymphoscintigraphy with triangulated body marking is a clinical option that can further reduce morbidity, more than when a hand held gamma probe alone is utilized. Unfortunately it is often either overlooked or not fully understood, and thus not utilized. This results in the unnecessary loss of an opportunity to further reduce morbidity.
Optimized lymphoscintigraphy and triangulated body marking provides a detailed 3 dimensional map of the number and location of the sentinel nodes, available before the first incision is made. The number, location, relevance based on time/sequence of appearance of the nodes, all can influence 1) where the incision is made, 2) how extensive the dissection is, and 3) how many nodes are removed. In addition, complex patterns can arise from injections. These include prominent lymphatic channels, pseudo-sentinel nodes, echelon and reverse echelon nodes and even contamination, which are much more difficult to access with the probe only. With the detailed information provided by optimized lymphoscintigraphy and triangulated body marking, the surgeon can approach the axilla in a more enlightened fashion, in contrast to when the less informed probe only method is used. This allows for better planning, resulting in the best cosmetic effect and less trauma to the tissues, further reducing morbidity while maintaining adequate sampling of the sentinel node(s).
==== Body
Introduction
The only goal of sentinel lymph node biopsy (SLNB) is to prevent/reduce morbidity associated with axillary lymph node dissection (ALND) while maintaining or enhancing sensitivity for detecting nodal disease. It has been extensively demonstrated that SLNB is associated with lower morbidity than ALND [1-24]. Lymphoscintigraphy potentially offers further reductions in morbidity, as compared to a probe-only methodology.
Some surgeons do not utilize lymphoscintigraphy at the time of SLNB and controversies continue [11,25-36]. One factor that adds confusion to understanding the value of lymphoscintigraphy is that when it is performed, there is often wide variation in techniques and quality [37-40], (figure 1). However, when optimized lymphoscintigraphy with triangulated patient body marking is not performed, an opportunity to further reduce morbidity is missed [38,40,41], [figure 2, table 1, 2, 3]. Below is a discussion of the advantages of lymphoscintigraphy, in which we seek to clarify misconceptions about its utility, discuss optimal techniques, provide an updated review of literature available on the added benefits of lymphoscintigraphy, and discuss its value in special situations.
Figure 1 (A*) TOP: Labeled as a "typical" lymphoscintigraphy finding in an article disputing the value of lymphoscintigraphy, the lateral view depicts the poor quality of injection and imaging technique [37]. The injection site is represented by the solid arrow, the faint, barely visible SN by the open arrow. (B) BOTTOM: Right lateral view of typical/average result from optimized injection and imaging protocol showing injection sites (solid arrow) and bright sentinel node (open arrow) as well as lymphatic channel leading to sentinel node [38]. In many cases, even much brighter nodes than depicted in B are found. *Reprinted from Am J Surg. 177, Burak WE Jr, Routine preoperative lymphoscintigraphy is not necessary prior to sentinel node biopsy for breast cancer, 445–449., 1999, with permission from Elsevier Ltd.; Excerpta Medica Inc . [37].
Figure 2 Schematic of triangulated patient body marking technique. Different colored permanent markers are used to place reference points on the patient's body corresponding to the location of a sentinel node along a particular projection. With this form of triangulation, the location of the sentinel nodes can be defined in 3 dimensions along appropriate triangulation lines. The arm is maintained in the surgical position (90°) to eliminate shifting of skin markings*. The rotation of the torso referenced to the floor must be kept constant during both imaging and surgery for the relationships to remain valid, or compensated for by equally shifted projections if rotation is desired during surgery [40]. *Adapted, revised and used with permission from Radiographics . 2004;24:121–145. Krynyckyi BR, et al. RSNA Publications, Oak Brook, IL. [ref. 38].
Table 1 Comparisons of average chronic pain and numbness/paresthesia morbidity between LS groups (+) performing lymphoscintigraphy and non LS groups (-) not performing lymphoscintigraphy in patients undergoing SLNB using radiotracer or using only dye. In general, studies using lymphoscintigraphy have much lower levels of chronic sensory morbidity. Data from original reference by Kim SC et al. [41].
Lymphoscintigraphy (+) Performed Lymphoscintigraphy (-) Not Performed
*Morbidity (Mor) Mor (%) Total Pt (N) References Mor (%) Total Pt (N) References #p-value
Pain (>9m) 13.77% 1365 1,2,4,9,10,11,14,20 28.67% 143 6 < 0.0001
Numbness/Paresthesia (>9m) 12.56% 677 1,4,9,11,13,14,17,20 23.14% 229 3,6,18 0.0003
*Adapted, revised, and used with permission from Kim SC et al: Using the intraoperative hand held probe without lymphoscintigraphy or using only dye correlates with higher sensory morbidity following sentinel lymph node biopsy in breast cancer: A review of the literature. World J Surg Oncol 2005, 3:64. [41].
#Statistics used to generate the p value: Fisher's exact test (2-tailed). A result was considered to be significant only if the p-value was lower than 0.05.
Table 2 Updated comparisons of average chronic pain and numbness/paresthesia morbidity between LS groups (+) performing lymphoscintigraphy and non LS groups (-) not performing lymphoscintigraphy in patients undergoing SLNB using radiotracer or using only dye. In general, studies using lymphoscintigraphy continue to have much lower levels of chronic sensory morbidity. Updated data by incorporation of four new references [21,22,23,24].
Lymphoscintigraphy (+) Performed Lymphoscintigraphy (-) Not Performed
*Morbidity (Mor) Mor (%) Total Pt (N) References Mor (%) Total Pt (N) References #p-value
Pain (>9m) 14.32% 1508 1,2,4,9,10,11,14,20,22,24 28.67% 143 6 < 0.0001
Numbness/Paresthesia (>9m) 9.22% 1052 1,4,9,11,13,14,17,20,23,24 23.17% 315 3,6,18,21t < 0.0001
*Adapted, revised, updated and used with permission from Kim SC et al: Using the intraoperative hand held probe without lymphoscintigraphy or using only dye correlates with higher sensory morbidity following sentinel lymph node biopsy in breast cancer: A review of the literature. World J Surg Oncol 2005, 3:64. [41].
tUse of LS is variable/suboptimal [21]. See text for details.
#Statistics used to generate the p value: Fisher's exact test (2-tailed). A result was considered to be significant only if the p-value was lower than 0.05.
Table 3 Potential advantages of optimized lymphoscintigraphy
- Reduces morbidity compared to probe only method.
- Reduces short and long term costs of treatment resulting from morbidity.
- Facilitates minimal invasiveness, improving cosmetic results.
- Assists surgeons in planning their approach and harvesting the SN.
- Provides additional guidance for surgeons who are learning the SLNB technique.
- Reduces overall surgical costs (anesthesia time, operating room utilization) by shortening surgery.
- Provides a wide field of view survey covering multiple lymph node basins simultaneously improving staging.
- Diffusion fields emanating from injection sites are defined and assessed for partly hidden nodes by scaling.
- The effects of breast displacement maneuvers are easily assessed.
- Delineates multiple SN nodes, their position and intensity along lymphatic channels, time of appearance.
- Delineates intervening nodes in unexpected positions when prominent lymphatic channels are present.
- Estimates the position of the SNs in the body from triangulated body marking (TBM).
- Surface contamination and other quality control issues are easy to detect and implement.
- Dynamic imaging is possible and its potential benefits in select cases.
- Assesses the intensity of the SN for next day surgery and determines the need for additional injections.
- Alerts to a failed node visualization (tumor replacement) and the possibility of a more extensive ALND.
- Delineates reverse echelon nodes, persistent lymphatic pools/dilations, end-on effects.
- Guides the planning of radiation ports with the inclusion of internal mammary chains when present.
- Sitting views can resolve clumped nodes, not possible with the probe after anesthesia.
- Improved localizing performance in obese patients compared to the probe before incision.
- Reduces the chance of un-harvested SNs (false negatives) through comparisons (numerical/positional/intensity) with images.
Potential disadvantages of lymphoscintigraphy
- Additional cost of procedure.
- Technically effort-intensive to fully optimize.
- Can delay surgery if not scheduled appropriately.
- Additional patient time required and any associated discomfort during imaging.
- Inadequate reimbursement for those performing it.
Advantages of lymphoscintigraphy
Wide field of view
Optimized lymphoscintigraphy provides detailed, comprehensive information on sentinel node (SN) location and on drainage routes. It has the advantage of simultaneously sampling the entire chest for minutes at a time, in contrast to the probe, which samples only one particular point for a short time [38,42-45]. Given this enhanced global sensitivity, the camera is superior in the initial survey. In addition, the images allow identification of internal mammary and other extra axillary nodes [46-48] and rare cases of drainage to the contralateral axilla [49-53]. When SLNB is performed in patients with previous breast implant augmentation (where complex patterns of activity can arise), or in cases of previous SLNB, the images can prove invaluable [54-57].
Unique drainage patterns visualized
In melanoma, lymphoscintigraphy is accepted because globally unique drainage patterns exist. Information from the images potentially improves staging by mapping out drainage basins [58,59]. This advantage can be applied to the axilla in breast cancer patients, but on a smaller, finer scale. Even limited to the region of the axilla, lymphoscintigraphy can provide similar, detailed information, as the axilla contains multiple nodes at various levels.
Triangulated body markings improve precision
Utilizing lymphoscintigraphy, an accurate initial incision location and dissection route can be chosen in advance. When the patient's skin is marked with triangulation points during lymphoscintigraphy, the surgeon can place the incision optimally. The triangulation marks result in a 3-Dimensional pattern, which can help to direct the surgeon to the location of the nodes, especially in obese patients or those with faint SNs [38,40,43], (Figure 2). Consequently, the initial incision location and dissection route can be better planned for the best cosmetics and least tissue disruption. This is in contrast to an approach using only the probe without surface markings, which might be equivocal during the initial survey in some patients.
Invasiveness can be further minimized
Clinical practice reveals a clear trend for minimally invasive techniques and breast conservation during both initial diagnosis, staging and subsequent surgical treatment [60-66]. The additional localizing information provided by optimized lymphoscintigraphy will allow for fine tuning of the axillary surgical approach in many patients, further optimizing morbidity reduction.
Additional views (standing/sitting) are possible
The standing/sitting views that are possible with lymphoscintigraphy further improve accuracy. These views reveal adjacent nodes hidden by the injection site scatter. They also resolve "clumping" of sentinel nodes that can occur in the axilla in the supine patient while also eliminating negative "end on effects" of lymphatic channels which will be discussed below [38,43,67-69]. This is important in delineating the true number of radioactive nodes, and also informs the surgeons of what to expect, and how many nodes to potentially remove [70,71]. The standing position obviously can not be performed with the probe at the initial survey before incision once anesthesia is administered.
Optimizing techniques
Injection technique and Hot nodes
Injection technique is an important factor contributing to the ease of finding the SN and the success of SLNB. Areolar-cutaneous junction injections and similar injections under the nipple increase SN activity, making the SN easier to find with the probe while generating optimal images [38,43,72]. These injections are very efficient in delivering activity to the SN, more so than perilesional or even intra/sub-dermal injections [72]. This is particularly important in the obese patient, where fat attenuates radioactivity and also increases distances between the SN and probe. In the setting of 15 cm of fat, less than 20% of the signal is left after attenuation. Increased distance further weakens the signal by 1/d2 (d = distance from probe to node) and directionality suffers [42,73]. Gamma camera sensitivity, conversely, does not appreciably change over distance [73].
Augmenting activity in the SN also facilitates next day surgery protocols [11,38,43,72,74-76]. Eighteen hours after injection, the 99mTc radiotracer has decayed to where only 12.5% of the original activity remains in the SN [73]. Performing injections and obtaining images using protocols that augment SN activity on the day before surgery, will alleviate surgical scheduling issues/delays. It can also save operating room time, by avoiding potential delays caused by starting the technique in the morning. These advantages will result in a cost savings [74-76].
Role of blue dye
When fully optimized lymphoscintigraphy with hybrid combination radiotracer injections is performed [38,43,72,77], it is no longer necessary to utilize blue dye as a primary method of finding the SN. In this setting, blue dye serves primarily as a backup (in the rare cases when a radioactive node is not detected) or as a secondary method to find the SN. Dye also serves as a potential visual guide when probe directionality is occasionally poor. Overall, dye provides less benefit than radiotracer as noted in several studies [78-80].
An exclusively blue dye technique can be viewed as not fully fulfilling the primary goal of morbidity reduction that SLNB promises. It necessitates more extensive dissection, as the lymphatic ducts leading to nodes are exposed until the SNs are found. In comparison, probe guided SN extraction can variably detect the SN directly through tissues, and further guidance is provided by lymphoscintigraphy images and skin markings [81-83].
Some early studies have shown 19.7% to 32.2% of SNs detected by dye alone in patients where radiotracer was also used [84,85]. However, these studies utilized inefficient perilesional injection techniques and no imaging methods (probe guided only).
In contrast, King et al. used dermal injections of radiocolloid employing lymphoscintigraphy and perilesional injections of dye. It was demonstrated that in 1719 procedures, only 1.9% of all the SNs were blue-only, and did not contain radioactivity detected by probe [86]. In a subgroup of procedures where smaller volumes of dye were used (0.1 ml–1.0 ml), only 1.3% of SNs were identified by blue dye alone [86]. The rate of blue only SNs positive for disease was higher however, at 10.5%. This may reflect the lack of simultaneous perilesional and areolar radiotracer injections as part of a hybrid injection technique as suggested by our group, as only 85.8% of studies demonstrated nodes on the images [72,86]. With experience in using radiotracer, use of dye becomes less relevant as was demonstrated in a study of 500 patients by Derossis et al. where a SN identified only with dye and containing disease was seen in only 2% of cases [87].
In a recent study, Degnim et al. [88] report on 418 cases in which radiotracer and dye were concurrently administered for SLNB. In 380 of these, SNs were identified on the lymphoscintigraphy images, and were recovered with the probe and/or by visual guidance from the blue dye. In only 3 of these 380 cases (0.79%) was disease that altered patient staging found in a node that contained blue dye but did not contain radioactivity [88]. In other words, had dye not been used, disease that changed staging would have been missed in 0.79% of total cases where SNs were identified on lymphoscintigraphy [88]. It is possible that technical issues, i.e., suboptimal SN intensity resulting from using mainly perilesional and dermal over the tumor injections, played a role in these findings. SNs were demonstrated on lymphoscintigraphy images in only 90.9% (380/418) of total cases administered radiotracer and dye. This is generally below the expected SN lymphoscintigraphy visualization rate obtained with current optimal areolar radiotracer injection methods and camera imaging protocols. Optimal delivery of radiotracer to SNs is an important issue, because the SNs involved with tumor may have limited numbers of macrophages, which can make the nodes appear faint if the delivery of radiotracer is suboptimal [38,40,72,74]. Higher efficiency areolar injections will generally increase activity in SNs proportionally. This will reduce occurrences of faint, disease containing nodes that are below the threshold of detection by the probe and camera, resulting in these nodes being missed by these methods, but visualized with dye. Thus, it can be postulated that even lower rates of blue node only disease would have been found if more efficient methods of radiotracer injection and imaging were employed in this study. This needs to be formally investigated.
In a study utilizing a variant of our previously reported hybrid combination injections of dermal and perilesional radiocolloid [77,89], without using any blue dye injections at all, Freezer et al. demonstrated a total 2% false negative rate for SLNB in cases that received ALND who had dynamic lymphoscintigraphy with triangulated skin markings [90].
Lastly, King et al. [86] also reported on 143 patients undergoing prophylactic mastectomy and SN biopsy. Of these patients, 9.1% had occult carcinoma identified, and the SN was positive in only 1.4% (2/143). In these particular low risk patients, the estimated probability of a blue-only SN not containing any radioactivity that would alter patient stage is a very small fraction of that 1.4%. Thus the use of dye in any capacity in this limited context that may lead to additional dissection can certainly be questioned.
Allergic/anaphylactic reactions, tattooing of skin with surface injections and localized tissue inflammation have been reported as complications to dye [86,90,91].
Examination of arguments against the use of lymphoscintigraphy
Several articles on lymphoscintigraphy question the value of imaging, and report no significant difference in the number of SNs found with and without images [37,92,93]. However, it is important to note that both injection and imaging techniques were key factors in these reports. Universally, less effective perilesional injections and non-optimized imaging techniques were utilized.
McMasters et al. reported that only 56% of patients who had lymphoscintigraphy performed showed axillary nodes. In addition, 36.2% showed no drainage to any basin at all on the lymphoscintigraphy images [92]. However, these results were certainly influenced by technique. The study was multi-center, and was performed in 1997–1999, with only delayed views after 45–60 minutes (no dynamic images). Only perilesional injections were utilized, not the currently preferred areolar or dermal injections [92]. Considering the variation in quality of the lymphoscintigraphy procedures performed at the different centers, this study cannot accurately reflect what is possible with optimized techniques of imaging, injection and triangulated patient body marking [38,43-45,72,94-96]. In fact, with current techniques, rates of visualization on lymphoscintigraphy are 97%–100% [11,94-96].
In a study by Burak et al. the authors conclude that routine preoperative lymphoscintigraphy is not necessary. However, in this study, image quality was severely compromised, and sentinel nodes were noted on the images in only 70.8% of patients. In addition to poor image quality, this study was based on a relatively small sample size: 24 patients, (figure 1) [37]. Furthermore, imaging was performed utilizing a lead shielding technique, and the results were very poor when the tumors were in the upper outer quadrant.
In a larger, Department of Defense study by DuPont et al. involving 516 patients who had imaging in 1997–1999, only 65% of patients demonstrated axillary nodes during lymphoscintigraphy [93]. However, here technical aspects again affected the findings. In this study, a suboptimally narrow energy window of 10% was used as opposed to an upwardly offset 16%–18% energy window which is recommended [38,43,73,96]. Again, only perilesional injections were used, and shielding of the injection site was also employed. The reported methods of Burak et al. and DuPont et al. also suggest that it is likely that suboptimal collimators were used, along with un-optimized image acquisition energy settings [38,43,73,96,97]. These technical shortcomings can be redressed by using high quality cast (non-foil) collimators and optimal camera energy settings, which completely obviate the need for lead shielding. Lead shielding complicates lymphoscintigraphy imaging greatly, and can lead to artifacts, missed sentinel nodes and false nodes [38,96,97].
Clearly, these studies were influenced by significant technical limitations, as evidenced by the described methods and/or poor quality sample images presented (figure 1). Furthermore, suboptimal injection technique also contributed to the findings, in that the more efficient injection methods (dermal or areolar) were not utilized. Unfortunately, these articles [37,92,93] still continue to be cited as proof of the lack of usefulness of lymphoscintigraphy [98,99]. Most importantly, none of them address the issues of potential morbidity reduction that the newer imaging techniques provide, focusing instead mainly on the SN detection rate or false negative rate [37,40,41,92,93].
A new look at the literature on morbidity reduction with lymphoscintigraphy
Unfortunately, to date, no landmark study has been performed that directly compares the additional reduction in morbidity achieved by properly performed lymphoscintigraphy, (with optimal injection technique and triangulated patient body markings) vs. use of the probe alone or dye alone. In the absence of such a study, we recently performed a focused review of the literature, which indirectly sheds light on this question [41].
Kim SC et al. [41] reviewed 20 articles addressing the differences in morbidity between SLNB and ALND. Of these 20 articles, the authors identified 10 articles in which lymphoscintigraphy was used that were suitable for comparison to the 3 articles identified in which lymphoscintigraphy was not utilized [41].
The percentage of patients experiencing chronic sensory morbidity after SLNB in Kim's analysis of these articles are depicted in Table 1. It is very important to note that multiple confounding issues may exist when comparing morbidity findings among different studies. However, at the very least, a clear trend seems to be present in Kim's analysis, which reveals that nearly twice the chronic sensory morbidity was reported from SLNB in the articles not utilizing lymphoscintigraphy but using only the probe or using only dye, vs. those articles using both lymphoscintigraphy and the probe [41], [Table 1].
Subsequent to Kim's analysis, Purushotham AD et al. [21] reported on the results of a randomized controlled clinical trial conducted in patients with primary breast cancer, which sought to compare morbidity associated with SLNB and ALND. Three hospitals participated in this study, in which patients were randomly assigned to ALND or SLNB. The SLNB procedure included radiotracer and blue dye. Personal communication with authors of this study revealed the following: in all 3 hospitals, in the SLNB arm, in the majority of cases, lymphoscintigraphy was not performed and/or if performed, was done suboptimally. In one hospital, the use of lymphoscintigraphy was abandoned half way through the study due to a perceived lack of usefulness. In another hospital lymphoscintigraphy was not used in the vast majority of cases. In the third hospital, lymphoscintigraphy was used only because it was required by the clinical protocols of that hospital (a surgical teaching hospital), however, the imaging specialist's reports concerning the results of lymphoscintigraphy were not communicated to the surgeons prior to surgery [21]. In addition, at this hospital, SNs were not marked on the patient's bodies, thus the opportunity to utilize triangulated body marking, which provides surgeons with a reference for SN location in the body, was missed [21,38,40,44,45,57].
Updated evaluation of the literature
In the vast majority of the cases in the study by Purushotham AD et al., lymphoscintigraphy was either used suboptimally or not at all [21]. Therefore, for the purposes of comparative analysis to other studies, it can be considered to be a non-lymphoscintigraphy study. Purushotham's study (non lymphoscintigraphy) [21], along with three new studies in which lymphoscintigraphy was used [22-24], was subjected to the inclusion/exclusion criteria utilized by Kim et al. [41] in their previous analysis, and the data from these four new studies was integrated into Kim's previous data. A new analysis of the updated dataset was performed. The results are represented in Table 2.
This table shows that the incorporation of the new data confirms and strengthens the trend suggested in Kim's previous analysis [41]. The updated analysis demonstrates over twice the chronic sensory morbidity among the studies not using lymphoscintigraphy but using probe-only or dye-only (P < 0.0001.)
It is interesting to note that, in general, the authors of the articles with the highest long term sensory morbidity who did not use lymphoscintigraphy, or abandoned/discounted (as shown in Table 1 and 2 above) have also published the vast majority of articles questioning its overall value [3,6,18,21,37,92,98-105]. This may be due to their past experiences with lymphoscintigraphy, which have convinced them of its lack of utility. However, it is quite possible that the lymphoscintigraphy protocols that were utilized in the past by these authors may have been affected by serious technical issues (such as were described above), resulting in suboptimal images of limited utility. Furthermore, it is likely that triangulated body marking was not used to guide and reduce dissection and subsequent morbidity. Based on our extensive clinical experience, we are aware that there is clearly great variation in the quality and methods of lymphoscintigraphy being practiced [37-40].
Intraoperative injections vs. lymphoscintigraphy
Layeeque et al. propose injecting the 99mTc sulfur colloid intra-operatively to eliminate the pain of injection, suggesting that vasovagal episodes and pain occur 10%–20% of the time with preoperative radiotracer injections [106]. In our experience pain can be well controlled by a combination of topical anesthetic applied to the skin and the simultaneous addition of anesthetic to the 99mTc sulfur colloid syringe [38,43,72,77].
Layeeque et al. [106] also suggests that with intraoperative areolar radiotracer injections, lymphoscintigraphy can be avoided as a result of the improved efficiency of delivering radiocolloid to the SN provided by surface injections and by the rapid flow of tracer to the SN from the injection site, which can reach the sentinel node before the injection is completed [38,43,72,107]. Hotter nodes are easier to find with the probe, and there have been minor advances in probe design, promising slightly better directionality in future models [108]. However, these superficial injection techniques are accompanied by unique features that make the images obtained during lymphoscintigraphy, including delayed views, all the more important.
Performed during surgery or in the nuclear medicine department, areolar injections will produce very prominent lymphatic channels. These can complicate the removal of nodes due to the extreme activity that can be present in them. Immediately after areolar injection, a very dynamic process occurs. Prominent channels appear, often multiple, that often course a tortuous path [38,43-45,69,72,96,107,109-111]. Nodes can blend in with channels for over 30–120 minutes after injection. Additionally, channels can track superficially above the axillary SN before coursing internally and inferiorly to the SN, in an inverted J pattern [38,43,72]. At inflections in the channels, activity can appear as foci, since the observer is looking at times down/through the length of the channel as opposed to perpendicular to it (the end on effect) [38,43,69,77,107]. These end on effects are also noted in all regions of the breast. Dilations/ectasias, which appear immediately after injection, pose an additional problem for the surgeon as they represent pseudo sentinel nodes and can appear as distinct foci. These are much more common with the areolar injection techniques compared to perilesional injections, as much more activity over a shorter time period is concentrated in the lymphatic channels [38,43,72,107].
During the initial 30–120 minutes after intraoperative injection, the surgeon is faced with a constantly changing pattern of radioactivity. Pursuing what appear to be foci that in fact represent end on effects or pseudo sentinel nodes and not real nodes, can result in unnecessary dissection when using only the probe [38,43,69,72,77,107,111]. Because these complex patterns arise immediately after injection and the changes continue over time, the information the camera provides in the form of dynamic images, multi-angle views and delayed views, is valuable in resolving the true nature of the patterns, a process in which the probe is severely disadvantaged. In fact, intraoperative injections could actually prolong surgery and increase dissection/morbidity in some patients, as a result of the complex post injection dynamic patterns described above [38,43,45,55,68,69,72,74,96,106,107,110-112]. If internal mammary sentinel nodes are deemed important to visualize, then the use of concurrent perilesional injections as part of a hybrid injection protocol of areolar and perilesional injections is necessary, as areolar injections do not delineate internal mammary nodes to any extent [38,72,113-115]. Perilesional injections require much more time to visualize the sentinel nodes than areolar injections.
Value of lymphoscintigraphy in special situations
There are a number of special situations in which information provided by lymphoscintigraphy is very valuable. These include situations of non-visualization of nodes, training, contamination, free pertechnetate, pregnancy and the elderly.
Even with the best areolar injection techniques, there are rare occasions when the nodes are not seen or only appear after an extended time period [72]. This can occur in older patients, obese patients or those with prior lumpectomies [43,72,77,110,116]. Knowing that the activity is weak or absent, an additional injection can be performed employing a higher dose and/or volume of radiotracer [40,72,112].
A final quality control check of the radiotracer occurs with imaging. This will readily show patterns of free pertechnetate, as well as surface contamination, which are more difficult to detect with the probe [38,40,77].
In situations where surgeons are in the initial steps of learning SLNB, a lack of lymphoscintigraphy images and triangulation reference points can be detrimental to the patient. SLNB is being performed with increasing frequency, fueled by demand from patients and prevailing trends. In the past SLNB was performed mostly by experienced investigators with a strong commitment to the technique. However, with rising demand, a greater number of less experienced mainstream surgeons are adopting SLNB (in some cases reluctantly), and performing the procedure. Here the lymphoscintigraphy images serve as a vital training tool, and can support those surgeons who are newly learning the technique of SLNB. Simulators have been also developed that can assist, along with the images, in training surgeons [117,118].
In pregnant women and in the elderly, SLNB is safe and accurate [119-121]. Since time spent under anesthesia in these patients should be minimized, knowing the number and location of sentinel nodes and lymphatic tracts before surgery will expedite SLNB removal, and accomplish the goal of minimizing anesthesia time.
Limiting the numbers of nodes removed
Using lymphoscintigraphy and triangulation, Kennedy et al. demonstrated that little benefit in additional sensitivity results from removing more than two sentinel nodes [122]. Similarly using lymphoscintigraphy, Schrenk et al. suggest that excising more than three nodes adds little to accuracy [123]. Identifying which node appears first and finding the location of the subsequent, more distant echelon nodes is even more important with areolar injections than perilesional injections. This is because areolar injections tend to delineate a greater number of echelon nodes, as a greater percentage of the injected activity enters more directly into the lymphatic channels and is available to spill down over to distant echelon nodes [38,40,43,72,74].
If faced with 6 hot nodes, it is crucial to know which nodes are more important to excise. Only the first two to four need to be excised if the sequence of appearance and position along the lymphatic chain is known, especially if both perilesional and areolar injections of radiotracer first drain to the same primary SN. Therefore, removal of all remote echelon nodes is probably not warranted in select patients with a very low probability of nodal disease. Surgeons not utilizing lymphoscintigraphy will be faced with several dilemmas: 1) not knowing which node was drained to by single or combination injections, 2) not knowing which node appeared first and second along the lymphatic channel, 3) not knowing the relative position of all the nodes along the lymphatic channel and 4) not knowing their relative intensities to each other before incision. They will need to consider taking all nodes out as the hottest node is not necessarily the one with disease [124,125]. In contrast, when lymphoscintigraphy has been performed, surgeons can approach the axilla in a more informed fashion. Furthermore, if only a single node is seen on the images, and clearly drained to by both perilesional and areolar injections, then extensive dissection for additional nodes can be avoided to minimize morbidity.
Morbidity reduction is a central goal of SLNB
When nodes are "hot", any reasonably good surgeon can achieve good sensitivity with varying levels of dissection. The aim is to accomplish this with as little dissection as possible, and at the same time maintain or improve the sensitivity. Knowing in advance the total number, location and pattern of SNs and lymphatic channels will result in a reduction in the total time of surgery, anesthesia, operating room associated costs, in addition to improved sensitivity. Most importantly a more targeted surgical approach will result in a reduction in patient morbidity. We observed unique patterns of drainage that can have an impact on the false negative rate and morbidity [38,43,50,55,69-72,74,77,96,107,113,115]. Based on the vast experience in imaging of multiple authors as well as ourselves [11,38,40,42-45,47,48,50,55,57,69-72,74,77,96,107,111-115,126-129], the general techniques described here form the basis of a logical algorithm for patient management.
The fine details of optimizations for lymphoscintigraphy are extensive and relevant mainly to the imaging specialist. Factors associated with optimizing the procedure include collimator design, gamma camera energy windows, radiopharmaceutical preparation, pain control, injection location and technique, dynamic and multi-angle camera views, patient arm and body positions, triangulated body marking, breast displacement maneuvers, outlining techniques, image display and printing parameters and a thorough communication with the surgeon regarding the findings. A condensed, referenced protocol is presented in table 4.
Table 4 Suggested optimized lymphoscintigraphy technique
Camera/Outlining:
- High resolution low energy cast (non-foil) collimator [38,96,97].
- 128 × 128 matrix-dynamic, 256 × 256 matrix-static.
- Upwardly offset 99mTc energy windows and separate 57Co energy windows (122 kev) [38,43,77,96].
- Decayed 57Co sheet source transmission outlining to limit exposure [38,96].
Injection:
- Anesthetic cream (EMLA) applied to injection sites for 30+ minutes [38,43,77].
- Hybrid radiotracer injection technique: Concurrent perilesional (2–4 ml biased away from the axilla) and areolar-cutaneous "junction" injections "LymphoBoost " (LB), (0.2–1.0 ml). Total dose: 150–400+ uCi 99mTc sulfur colloid for same day injections and surgery, 500–1000+ uCi for next day surgery. Higher LB volumes towards 1.0 ml tend to visualize nodes quicker and brighter but delineate more echelon nodes compared to lower volumes of 0.2 ml [38,40,43,72,74].
- High specific activity preparation, 100% filtered [130].
- Lidocain added to sulfur colloid syringe for additional pain control [38,43,77].
- Mild/short massage only [131].
- Deeper sub-lesional injections for internal mammary SN visualization if deemed important [113].
- Contamination control [77].
Acquisition sequences:
- Optional post perilesional injection views.
- Dynamic lateral 100 frame 10 second images during areolar-cutaneous "junction" injection "LymphoBoost" (LB) [38,43,72,74].
- Optional immediate post dynamic early static sitting/standing views (see below).
- Delayed supine anterior and oblique 45° views with the arm out in the 90° surgical position and lateral views with the arm up towards the head with triangulated body marking of anterior and oblique 45° views.
- 57Co sheet source transmission outlining of anterior and lateral views [38,43,96].
- Sitting/Standing views highly recommended (see below).
Additional optional maneuvers:
- Perform perilesional injection followed by 30 minute (or more) delayed views followed by LB injection (dynamic see above). Alternately delete perilesional injections altogether (only inject LB).
- Adaptive Injection Technique (AIT), re-inject different volumes of radiotracer based on imaging results [40],(data pending publication).
- Tape breast displacement for small breasts, for large/pendulous breasts use sitting views (see below) [38,43,96].
- Prone imaging [112], MOVA position [127], next day follow up views if two day.
- Avoid lead shielding the injection site [42,96,97].
Triangulated body marking:
- See figure 2, [38,40,43].
Sitting/Standing views:
- Highly recommended end of study anterior and lateral sitting/standing views with arm out in the 90° surgical position with chest pressed up against collimator (best resolution), two 1 minute frames each position to address motion if it occurs [38,40,43,67-71,96]. Works best in large breasted women.
Display:
- Adjustment of upper level, gamma curve, pre-display low level data enhancement (pre-scale/contrast/threshold) and appropriate image summation [96].
- Viewing dynamic sequences in cine mode [107].
Printing:
- Two sets of images for final supine views (marking views): with and without 57Co transmission scan (when performed) [38,96].
- Print images large enough for surgeons to clearly see anatomy. Optionally print sitting views and/or dynamic sequences if important [38,96,107].
Reporting:
- Timely and detailed communications with surgeon before surgery to discuss findings, meaning/convention of markings and complex patterns. Number of SN based on supine and standing views, appearance sequence and perceived intensity, 3-D position in body, any extra-axillary or intramammary nodes, dilations/ectasias.
Summary
SLNB has essentially become the standard of care irrespective of pending prospective data [132]. Besides the critical questions of false negative rates, equally important emphasis should be placed on further reducing morbidity by optimization of sentinel node excisional techniques. In order to accomplish this objective, SLNB methodology should include 1) detailed, optimized lymphoscintigraphy, 2) maneuvers to increase activity within the SN and 3) triangulated patient body marking. At the present time, in general, these methods are often either not used (alone or in combination), or if they are used, are done so in a suboptimal manner. A paradigm shift in departmental methods is needed to incorporate these valuable techniques, in order to meet the objectives of minimally invasive surgery, breast conservation and morbidity reduction. While it is may be true that all women who have SLNB do not benefit directly from lymphoscintigraphy images, in the patients where the images make a difference and reduce morbidity, a very meaningful improvement in patient care will have been achieved.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
BRK developed the initial concept and contributed to data analysis, design, revision and preparation of manuscript.
MKS contributed to data analysis, design, revision and preparation of manuscript.
SCK contributed to data analysis, design, revision and preparation of manuscript.
DWK contributed to data analysis
AT contributed to preparation, design and revision of manuscript.
RMM contributed to data analysis and preparation of manuscript.
CK contributed to data analysis, design, and preparation of manuscript.
==== Refs
Schrenk P Rieger R Shamiyeh A Wayand W Morbidity following sentinel lymph node biopsy versus axillary lymph node dissection for patients with breast carcinoma Cancer 2000 88 608 614 10649254 10.1002/(SICI)1097-0142(20000201)88:3<608::AID-CNCR17>3.0.CO;2-K
Roumen RM Kuijt GP Liem IH van Beek MW Treatment of 100 patients with sentinel node-negative breast cancer without further axillary dissection Br J Surg 2001 88 1639 1643 11736979 10.1046/j.0007-1323.2001.01935.x
Burak WE Hollenbeck ST Zervos EE Hock KL Kemp LC Young DC Sentinel lymph node biopsy results in less postoperative morbidity compared with axillary lymph node dissection for breast cancer Am J Surg 2002 183 23 27 11869698 10.1016/S0002-9610(01)00848-0
Haid A Koberle-Wuhrer R Knauer M Burtscher J Fritzsche H Peschina W Jasarevic Z Ammann M Hergan K Sturn H Zimmermann G Morbidity of breast cancer patients following complete axillary dissection or sentinel node biopsy only: a comparative evaluation Breast Cancer Res Treat 2002 73 31 36 12083629 10.1023/A:1015234318582
Temple LK Baron R Cody HS 3rdFey JV Thaler HT Borgen PI Heerdt AS Montgomery LL Petrek JA Van Zee KJ Sensory morbidity after sentinel lymph node biopsy and axillary dissection: a prospective study of 233 women Ann Surg Oncol 2002 9 654 662 12167579 10.1245/aso.2002.9.7.654
Swenson KK Nissen MJ Ceronsky C Swenson L Lee MW Tuttle TM Comparison of side effects between sentinel lymph node and axillary lymph node dissection for breast cancer Ann Surg Oncol 2002 9 745 753 12374657 10.1245/ASO.2002.02.007
Haid A Kuehn T Konstantiniuk P Koberle-Wuhrer R Knauer M Kreienberg R Zimmermann G Shoulder-arm morbidity following axillary dissection and sentinel node only biopsy for breast cancer Eur J Surg Oncol 2002 28 705 710 12431466 10.1053/ejso.2002.1327
Leidenius M Leppanen E Krogerus L von Smitten K Motion restriction and axillary web syndrome after sentinel node biopsy and axillary clearance in breast cancer Am J Surg 2003 185 127 130 12559441 10.1016/S0002-9610(02)01214-X
Schijven MP Vingerhoets AJ Rutten HJ Nieuwenhuijzen GA Roumen RM van Bussel ME Voogd AC Comparison of morbidity between axillary lymph node dissection and sentinel node biopsy Eur J Surg Oncol 2003 29 341 350 12711287 10.1053/ejso.2002.1385
Blanchard DK Donohue JH Reynolds C Grant CS Relapse and morbidity in patients undergoing sentinel lymph node biopsy alone or with axillary dissection for breast cancer Arch Surg 2003 138 482 487 discussion 487–488. 12742949 10.1001/archsurg.138.5.482
Veronesi U Paganelli G Viale G Luini A Zurrida S Galimberti V Intra M Veronesi P Robertson C Maisonneuve P Renne G De Cicco C De Lucia F Gennari R A randomized comparison of sentinel-node biopsy with routine axillary dissection in breast cancer N Engl J Med 2003 349 546 553 12904519 10.1056/NEJMoa012782
Rietman JS Dijkstra PU Geertzen JH Baas P De Vries J Dolsma W Groothoff JW Eisma WH Hoekstra HJ Short-term morbidity of the upper limb after sentinel lymph node biopsy or axillary lymph node dissection for Stage I or II breast carcinoma Cancer 2003 98 690 696 12910511 10.1002/cncr.11545
Peintinger F Reitsamer R Stranzl H Ralph G Comparison of quality of life and arm complaints after axillary lymph node dissection vs sentinel lymph node biopsy in breast cancer patients Br J Cancer 2003 89 648 652 12915872 10.1038/sj.bjc.6601150
Baron RH Fey JV Borgen PI Van Zee KJ Eighteen sensations after breast cancer surgery: a two-year comparison of sentinel lymph node biopsy and axillary lymph node dissection Oncol Nurs Forum 2004 31 691 698 15252425
Armer J Fu MR Wainstock JM Zagar E Jacobs LK Lymphedema following breast cancer treatment, including sentinel lymph node biopsy Lymphology 2004 37 73 91 15328760
Ronka RH Pamilo MS von Smitten KA Leidenius MH Breast lymphedema after breast conserving treatment Acta Oncol 2004 43 551 557 15370612 10.1080/02841860410014867
Rietman JS Dijkstra PU Geertzen JH Baas P de Vries J Dolsma WV Groothoff JW Eisma WH Hoekstra HJ Treatment-related upper limb morbidity 1 year after sentinel lymph node biopsy or axillary lymph node dissection for stage I or II breast cancer Ann Surg Oncol 2004 11 1018 1024 15525832 10.1245/ASO.2004.03.512
Langer S Guenther JM Haigh PI Difronzo LA Lymphatic mapping improves staging and reduces morbidity in women undergoing total mastectomy for breast carcinoma Am Surg 2004 70 881 885 15529842
Luini A Gatti G Zurrida S Galimberti V Paganelli G Naninato P Caldarella P Rotmensz N Winnikow E Viale G The sentinel lymph node biopsy under local anesthesia in breast carcinoma: experience of the European Institute of Oncology and impact on quality of life Breast Cancer Res Treat 2005 89 69 74 15666199 10.1007/s10549-004-1473-y
Ronka R von Smitten K Tasmuth T Leidenius M One-year morbidity after sentinel node biopsy and breast surgery Breast 2005 14 28 36 15695078 10.1016/j.breast.2004.09.010
Purushotham AD Upponi S Klevesath MB Bobrow L Millar K Myles JP Duffy SW Morbidity after sentinel lymph node biopsy in primary breast cancer: results from a randomized controlled trial J Clin Oncol 2005 23 4312 4321 15994144 10.1200/JCO.2005.03.228
Fleissig A Fallowfield LJ Langridge CI Johnson L Newcombe RG Dixon JM Kissin M Mansel RE Post-operative arm morbidity and quality of life. Results of the ALMANAC randomised trial comparing sentinel node biopsy with standard axillary treatment in the management of patients with early breast cancer Breast Cancer Res Treat 2005 DOI:10.1007/s10549-005-9025-7 16163445
Arnaud S Houvenaeghel G Moutardier V Butarelli M Martino M Tallet A Braud AC Jacquemier J Julian-Reynier C Brenot-Rossi I Patients' and surgeons' perspectives on axillary surgery for breast cancer Eur J Surg Oncol 2004 30 735 743 15296987 10.1016/j.ejso.2004.05.007
Barranger E Dubernard G Fleurence J Antoine M Darai E Uzan S Subjective morbidity and quality of life after sentinel node biopsy and axillary lymph node dissection for breast cancer J Surg Oncol 2005 92 17 22 16180226 10.1002/jso.20343
Krag D Ashikaga T The design of trials comparing sentinel-node surgery and axillary resection N Engl J Med 2003 349 603 605 12904526 10.1056/NEJMe038105
Badwe RA Thorat MA Parmar VV Sentinel-node biopsy in breast cancer [letter] N Engl J Med 2003 349 1968 1971 author reply 1968–1971. 14614174 10.1056/NEJM200311133492017
De Salvo GL Del Bianco P Zavagno G Sentinel-node biopsy in breast cancer [letter] N Engl J Med 2003 349 1968 1971 author reply 1968–1971. 14626250 10.1056/NEJM200311133492017
Munster AM Sentinel-node biopsy in breast cancer [letter] N Engl J Med 2003 349 1968 1971 author reply 1968–1971. 14626251 10.1056/NEJM200311133492017
McMasters KM Sentinel-node biopsy in breast cancer [letter] N Engl J Med 2003 349 1968 1971 author reply 1968–1971. 14626252 10.1056/NEJM200311133492017
Veronesi U Maisonneuve P Sentinel-node biopsy in breast cancer [letter] N Engl J Med 2003 349 1968 1971 author reply 1968–1971. 10.1056/NEJMoa012782
Krag D Ashikaga T Sentinel-node biopsy in breast cancer [letter] N Engl J Med 2003 349 1968 1971 author reply 1968–1971. 10.1056/NEJMe038105
Singh Ranger G Mokbel K The evolving role of sentinel lymph node biopsy for breast cancer Eur J Surg Oncol 2003 29 423 425 12798744 10.1016/S0748-7983(03)00038-6
Kern KA Achieving the lowest false-negative rate in peritumoral breast lymphatic mapping: the oncologic search for the Holy Grail Ann Surg Oncol 2003 10 486 487 12794011 10.1245/ASO.2003.04.006
Cody HS 3rd Sentinel lymph node biopsy for breast cancer: does anybody not need one? Ann Surg Oncol 2003 10 1131 1132 14654465 10.1245/ASO.2003.10.905
McMasters KM The eternally enigmatic axilla: further controversy about axillary lymph nodes in breast cancer Ann Surg Oncol 2003 10 1128 1130 14654464 10.1245/ASO.2003.10.917
Rutgers EJ Sentinel node procedure in breast carcinoma: a valid tool to omit unnecessary axillary treatment or even more? Eur J Cancer 2004 40 182 186 14728931 10.1016/j.ejca.2003.09.025
Burak WE JrWalker MJ Yee LD Kim JA Saha S Hinkle G Olsen JO Pozderac R Farrar WB Routine preoperative lymphoscintigraphy is not necessary prior to sentinel node biopsy for breast cancer Am J Surg 1999 177 445 449 10414690 10.1016/S0002-9610(99)00088-4
Krynyckyi BR Kim CK Goyenechea MR Chan PT Zhang ZY Machac J Clinical breast lymphoscintigraphy: optimal techniques for performing studies, image atlas and analysis of images Radiographics 2004 24 121 145 14730041
Pandey M Deo SVS Maharajan R Fallacies of preoperative lymphoscintigraphy in detecting sentinel node in breast cancer World J Surg Oncol 2005 3 31 15927059 10.1186/1477-7819-3-31
Krynyckyi BR Kim SC Kim CK Preoperative lymphoscintigraphy and triangulated patient body marking are important parts of the sentinel node process in breast cancer World J Surg Oncol 2005 3 56 16120218 10.1186/1477-7819-3-56
Kim SC Kim DW Moadel RM Kim CK Chatterjee S Shafir MK Travis A Machac J Krynyckyi BR Using the intraoperative hand held probe without lymphoscintigraphy or using only dye correlates with higher sensory morbidity following sentinel lymph node biopsy in breast cancer: A review of the literature World J Surg Oncol 2005 3 64 16194276 10.1186/1477-7819-3-64
Keshtgar MRS Waddington WA Lakhani SR Ell PJ Keshtgar MRS The Sentinel Node in Surgical Oncology 1999 Berlin, Heidelberg: Springer-Verlag
Krynyckyi BR Kim CK Shafir MK Mosci K Machac J Freeman LM Breast cancer and its management, the utility and technique of lymphoscintigraphy Nuclear Medicine Annual; Philadelphia 2003 Lippincott William & Wilkins 131 169
Mariani G Moresco L Viale G Villa G Bagnasco M Canavese G Buscombe J Strauss HW Paganelli G Radioguided sentinel lymph node biopsy in breast cancer surgery J Nucl Med 2001 42 1198 1215 11483681
Mariani G Erba P Villa G Gipponi M Manca G Boni G Buffoni F Castagnola F Paganelli G Strauss HW Lymphoscintigraphic and intraoperative detection of the sentinel lymph node in breast cancer patients: the nuclear medicine perspective J Surg Oncol 2004 85 112 122 14991882 10.1002/jso.20023
Bevilacqua JL Gucciardo G Cody HS MacDonald KA Sacchini V Borgen PI Van Zee KJ A selection algorithm for internal mammary sentinel lymph node biopsy in breast cancer Eur J Surg Oncol 2002 28 603 614 12359195 10.1053/ejso.2002.1269
Tanis PJ Nieweg OE Valdes Olmos RA Peterse JL Rutgers EJ Hoefnagel CA Kroon BB Impact of non-axillary sentinel node biopsy on staging and treatment of breast cancer patients Br J Cancer 2002 87 705 710 12232750 10.1038/sj.bjc.6600359
Estourgie SH Tanis PJ Nieweg OE Valdes Olmos RA Rutgers EJ Kroon BB Should the hunt for internal mammary chain sentinel nodes begin? An evaluation of 150 breast cancer patients Ann Surg Oncol 2003 10 935 941 14527914 10.1245/ASO.2003.02.015
Allweis TM Parson B Klein M Sklair-Levy M Maly B Rivkind A Uziely B Breast cancer draining to bilateral axillary sentinel lymph nodes Surgery 2003 134 506 508 14555941 10.1067/S0039-6060(03)00247-2
Lim I Shim J Goyenechea M Kim CK Krynyckyi BR Drainage across midline to sentinel nodes in the contralateral axilla in breast cancer Clin Nucl Med 2004 29 346 347 15166879 10.1097/01.rlu.0000127087.12686.81
Barranger E Montravers F Kerrou K Marpeau O Raileanu I Antoine M Talbot JN Uzan S Contralateral axillary sentinel lymph node drainage in breast cancer: A case report J Surg Oncol 2004 86 167 169 15170657 10.1002/jso.20056
Shimazu K Tamaki Y Taguchi T Akazawa K Inoue T Noguchi S Sentinel lymph node biopsy using periareolar injection of radiocolloid for patients with neoadjuvant chemotherapy-treated breast carcinoma Cancer 2004 100 2555 2561 15197796 10.1002/cncr.20242
Carmon M Mintz A Hain D Olsha O Clinical implications of contralateral axillary sentinel lymph nodes Breast 2005 DOI:10.1016/j.breast.2005.03.007 16005230
Gray RJ Forstner-Barthell AW Pockaj BA Schild SE Halyard MY Breast-conserving therapy and sentinel lymph node biopsy are feasible in cancer patients with previous implant breast augmentation Am J Surg 2004 188 122 125 15249236 10.1016/j.amjsurg.2003.12.062
Krynyckyi BR Shim J Goyenechea M Kim CK Layering of activity around a breast implant capsule during lymphoscintigraphy Clin Nucl Med 2004 29 598 599 15311140 10.1097/01.rlu.0000135002.34305.d6
Munhoz AM Aldrighi C Buschpiegel C Ono C Montag E Fells K Arruda E Sturtz G Kovac P Filassi JR Gemperli R Ferreira MC The feasibility of sentinel lymph node detection in patients with previous transaxillary implant breast augmentation: preliminary results Aesthetic Plast Surg 2005 29 163 168 15959693 10.1007/s00266-004-0103-8
Intra M Trifiro G Viale G Rotmensz N Gentilini OD Soteldo J Galimberti V Veronesi P Luini A Paganelli G Veronesi U Second biopsy of axillary sentinel lymph node for reappearing breast cancer after previous sentinel lymph node biopsy Ann Surg Oncol 2005 12 895 899 16195833 10.1245/ASO.2005.10.018
Morris KT Stevens JS Pommier RF Fletcher WS Vetto JT Usefulness of preoperative lymphoscintigraphy for the identification of sentinel lymph nodes in melanoma Am J Surg 2001 181 423 426 11448434 10.1016/S0002-9610(01)00604-3
Uren RF Howman-Giles R Thompson JF Patterns of lymphatic drainage from the skin in patients with melanoma J Nucl Med 2003 44 570 582 12679402
Paepke S Schwarz-Boeger U Kiechle M Jacobs VR Axillary dissection with access minimized (ADAM): a new technique for lymph node dissection in conservative surgery for breast cancer Int J Fertil Womens Med 2003 48 232 237 14626380
Ogawa Y Ishikawa T Sawada T Chung SH Osaka H Takashima T Onoda N Kato Y Ochi H Hirakawa K Thoracoscopic internal mammary sentinel node biopsy for breast cancer Surg Endosc 2003 17 315 319 12399844 10.1007/s00464-002-9042-0
Singletary SE Dowlatshahi K Dooley W Hollenbeck ST Kern K Kuerer H Newman LA Simmons R Whitworth P Minimally invasive operation for breast cancer Curr Probl Surg 2004 41 394 447 15114298 10.1016/j.cpsurg.2003.12.002
Yamamoto D Tanaka K A review of mammary ductoscopy in breast cancer Breast J 2004 10 295 297 15239786 10.1111/j.1075-122X.2004.21488.x
Veronesi U Cascinelli N Mariani L Greco M Saccozzi R Luini A Aguilar M Marubini E Twenty-year follow-up of a randomized study comparing breast conserving surgery with radical mastectomy for early breast cancer N Engl J Med 2002 347 1227 1232 12393819 10.1056/NEJMoa020989
Luini A Zurrida S Paganelli G Galimberti V Sacchini V Monti S Veronesi P Viale G Veronesi U Comparison of radioguided excision with wire localization of occult breast lesions Br J Surg 1999 86 522 525 10215829 10.1046/j.1365-2168.1999.01078.x
Krynyckyi BR Shafir M Zhang Z Kim CK Machac J TC-99M MAA improves on wire in localizing non-palpable breast lesions [abstract] J Nuc Med 2004 45 146P
Pierini A Dworkin HJ Is the upright position more sensitive than the supine position in breast cancer sentinel node lymphoscintigraphy? Clin Nucl Med 2001 26 823 825 11564917 10.1097/00003072-200110000-00003
Uren RF Howman-Giles R Renwick SB Gillett D Lymphatic mapping of the breast: locating the sentinel lymph nodes World J Surg 2001 25 789 793 11376417 10.1007/s00268-001-0006-7
Kim S Youssef I Kim CK Machac J Krynyckyi BR Prominent lymphatic channels simulating sentinel nodes: The utility of standing and delayed views in delineating the true number and position of nodes and the implications for further morbidity reduction Clin Nucl Med 2005 30 794 796 16319635 10.1097/01.rlu.0000187496.54900.a2
Kim SH Kim SC Kim DW Kim YJ Youssef IM Kim CK Machac J Krynyckyi BR Can different arm and body positions help in detecting more sentinel lymph nodes (SN) during lymphoscintigraphy (LS) [abstract]? J Nucl Med 2005 46 405P 15750151
Kim SH Kim SC Kim DW Machac J Kim CK Krynyckyi BR The effect of different arm positions on sentinel node localization during lymphoscintigraphy [abstract] J Nucl Med 2005 46 405P 15750151
Krynyckyi BR Kim CK Mosci K Fedorciw BJ Zhang ZY Lipszyc H Machac J Areolar-cutaneous "junction" injections to augment sentinel node count activity Clin Nucl Med 2003 28 97 107 12544125 10.1097/00003072-200302000-00002
Cherry SR Sorenson JA Phelps ME Physics in Nuclear Medicine 2003 third WB Saunders, Philadelphia, PA 31 43 65–88; 165–183.
Kim S Kim CK Krynyckyi BR Areolar-cutaneous junction injection boosts activity in sentinel node by more than 50 times compared to perilesional injection: implications for morbidity reduction Am Surg 2005 71 77 80 15757063
Yeung HW Cody III HS Turlakow A Riedel ER Fey J Gonen M Nunez R Yeh SD Larson SM Lymphoscintigraphy and sentinel node localization in breast cancer patients: a comparison between 1-day and 2-day protocols J Nucl Med 2001 42 420 423 11337517
Solorzano CC Ross MI Delpassand E Mirza N Akins JS Kuerer HM Meric F Ames FC Newman L Feig B Singletary SE Hunt KK Utility of breast sentinel lymph node biopsy using day-before-surgery injection of high-dose 99mTc-labeled sulfur colloid Ann Surg Oncol 2001 8 821 827 11776497
Krynyckyi BR Miner M Ragonese JM Firestone M Kim CK Machac J Technical aspects of performing lymphoscintigraphy: Optimization of methods used to obtain images Clin Nucl Med 2000 25 978 985 11129163 10.1097/00003072-200012000-00003
Kern KA Concordance and validation study of sentinel lymph node biopsy for breast cancer using subareolar injection of blue dye and technetium 99m sulfur colloid J Am Coll Surg 2002 195 467 475 12375751 10.1016/S1072-7515(02)01312-1
Lin KM Patel TH Ray A Ota M Jacobs L Kuvshinoff B Chung M Watson M Ota DM Intradermal radioisotope is superior to peritumoral blue dye or radioisotope in identifying breast cancer sentinel nodes J Am Coll Surg 2004 199 561 566 15454139 10.1016/j.jamcollsurg.2004.06.018
Marrazzo A Taormina P Noto A Cardinale G Casa L Mercadante S Lo Gerfo D David M Localization of the sentinel lymph node in breast cancer: prospective comparison of vital staining and radioactive tracing methods Chir Ital 2004 56 621 627 15553432
Giuliano AE Kirgan DM Guenther JM Morton DL Lymphatic mapping and sentinel lymphadenectomy for breast cancer Ann Surg 1994 220 391 398 8092905
Kern KA Sentinel lymph node mapping in breast cancer using subareolar injection of blue dye J Am Coll Surg 1999 189 539 545 10589589 10.1016/S1072-7515(99)00200-8
Canavese G Gipponi M Catturich A Di Somma C Vecchio C Rosato F Tomei D Cafiero F Moresco L Nicolo G Carli F Villa G Buffoni F Badellino F Sentinel lymph node mapping opens a new perspective in the surgical management of early-stage breast cancer: a combined approach with vital blue dye lymphatic mapping and radioguided surgery Semin Surg Oncol 1998 15 272 277 9829386 10.1002/(SICI)1098-2388(199812)15:4<272::AID-SSU17>3.0.CO;2-I
Cox CE Pendas S Cox JM Joseph E Shons AR Yeatman T Ku NN Lyman GH Berman C Haddad F Reintgen DS Guidelines for sentinel node biopsy and lymphatic mapping of patients with breast cancer Ann Surg 1998 227 645 651 9605656 10.1097/00000658-199805000-00005
Bass SS Lyman GH McCann CR Ku NN Berman C Durand K Bolano M Cox S Salud C Reintgen DS Cox CE Lymphatic mapping and sentinel lymph node biopsy Breast J 1999 5 288 295 11348304 10.1046/j.1524-4741.1999.00001.x
King TA Fey JV Van Zee KJ Heerdt AS Gemignani ML Port ER Sclafani L Sacchini V Petrek JA Cody HS 3rdBorgen PI Montgomery LL A prospective analysis of the effect of blue-dye volume on sentinel lymph node mapping success and incidence of allergic reaction in patients with breast cancer Ann Surg Oncol 2004 11 535 541 15123464 10.1245/ASO.2004.10.014
Derossis AM Fey J Yeung H Yeh SD Heerdt AS Petrek J VanZee KJ Montgomery LL Borgen PI Cody HS 3rd A trend analysis of the relative value of blue dye and isotope localization in 2,000 consecutive cases of sentinel node biopsy for breast cancer J Am Coll Surg 2001 193 473 478 11708502 10.1016/S1072-7515(01)01038-9
Degnim AC Oh K Cimmino VM Diehl KM Chang AE Newman LA Sabel MS Is blue dye indicated for sentinel lymph node biopsy in breast cancer patients with a positive lymphoscintigram? Ann Surg Oncol 2005 12 712 717 16079955 10.1245/ASO.2005.06.017
Krynyckyi BR Firestone M Eskandar Y Kim CK Machac J Dual method injection technique for breast lymphoscintigraphy to maximize visualization of sentinel nodes [abstract] J Nuc Med 2000 41 281P
Feezor RJ Kasraeian A Copeland EM 3rdSchell SR Hochwald SN Cendan J Drane W Mastin S Wilkinson E Lind DS Sequential dermal-peritumoral radiocolloid injection for sentinel node biopsy for breast cancer: the University of Florida experience Am Surg 2002 68 684 688 12206602
Singh-Ranger G Mokbel K Capsular contraction following immediate reconstructive surgery for breast cancer – An association with methylene blue dye Int Semin Surg Oncol 2004 1 3 15285809 10.1186/1477-7800-1-3
McMasters KM Wong SL Tuttle TM Carlson DJ Brown CM Dirk Noyes R Glaser RL Vennekotter DJ Turk PS Tate PS Sardi A Edwards MJ Preoperative lymphoscintigraphy for breast cancer does not improve the ability to identify axillary sentinel lymph nodes Ann Surg 2000 231 724 731 10767794 10.1097/00000658-200005000-00013
Dupont EL Kamath VJ Ramnath EM Shivers SC Cox C Berman C Leight GS JrRoss MI Blumencranz P Reintgen DS DOD Breast Lymphatic Mapping Trial Investigators The role of lymphoscintigraphy in the management of the patient with breast cancer Ann Surg Oncol 2001 8 354 360 11352310
Merson M Fenaroli P Gianatti A Virotta G Giuliano LG Bonasegale A Bambina S Pericotti S Guerra U Tondini C Sentinel node biopsy in the surgical management of breast cancer: experience in a general hospital with a dedicated surgical team Breast 2004 13 200 205 15177422 10.1016/j.breast.2004.01.007
Trifiro G Viale G Gentilini O Travaini LL Paganelli G Sentinel node detection in pre-operative axillary staging Eur J Nucl Med Mol Imaging 2004 31 S46 55 15103506 10.1007/s00259-004-1526-9
Krynyckyi BR Sata S Zolty I Kim CK Knesaurek K Reducing exposure from 57Co sources during breast lymphoscintigraphy by optimizing energy windows and other suggested enhancements of acquisition and the display of images J Nucl Med Technol 2004 32 198 205 15576341
Tsushima H Yamanaga T Shimonishi Y Ochi H Usefulness of imaging method without using lead plate for sentinel lymph node scintigraphy Kaku Igaku 2002 39 161 169 12058426
Kelley MC Hansen N McMasters KM Lymphatic mapping and sentinel lymphadenectomy for breast cancer Am J Surg 2004 188 49 61 15219485 10.1016/j.amjsurg.2003.10.028
Tuttle TM Technical advances in sentinel lymph node biopsy for breast cancer Am Surg 2004 70 407 413 15156948
Scoggins CR Chagpar AB Martin RC McMasters KM Should sentinel lymph-node biopsy be used routinely for staging melanoma and breast cancers? Nat Clin Pract Oncol 2005 2 448 455 16265014 10.1038/ncponc0293
Upponi SS McIntosh SA Wishart GC Balan KK Purushotham AD Sentinel lymph node biopsy in breast cancer – is lymphoscintigraphy really necessary? Eur J Surg Oncol 2002 28 479 480 12356023
Chagpar AB Kehdy F Scoggins CR Martin RC 2ndCarlson DJ Laidley AL El-Eid SE McGlothin TQ Noyes RD Ley PB Tuttle TM McMasters KM University of Louisville Breast Sentinel Lymph Node Study Effect of lymphoscintigraphy drainage patterns on sentinel lymph node biopsy in patients with breast cancer Am J Surg 2005 190 557 562 16164919 10.1016/j.amjsurg.2005.06.010
Guenther JM Collins JC Barnes G JrO'Connell TX Selective lymphoscintigraphy: a necessary adjunct to dye-directed sentinel node biopsy for breast cancer? Arch Surg 2000 135 1101 1105 10982518 10.1001/archsurg.135.9.1101
McIntosh SA Ravichandran D Balan KK Bobrow L Wishart GC Purushotham AD Sentinel lymph node biopsy in impalpable breast cancer Breast 2001 10 82 83 14965566 10.1054/brst.2000.0200
Purushotham AD Macmillan RD Wishart GC Advances in axillary surgery for breast cancer-time for a tailored approach Eur J Surg Oncol 2005 31 929 931 doi:10.1016/j.ejso.2005.07.008 16109469 10.1016/j.ejso.2005.07.008
Layeeque R Kepple J Henry-Tillman RS Adkins L Kass R Colvert M Gibson R Mancino A Korourian S Klimberg VS Intraoperative subareolar radioisotope injection for immediate sentinel lymph node biopsy Ann Surg 2004 239 841 848 15166963 10.1097/01.sla.0000128304.13522.00
Kim SH Shim J Kim CK Machac J Krynyckyi BR Reverse echelon node and a lymphatic ectasia in the same patient during breast lymphoscintigraphy: The importance of injection and imaging technique Br J Radiol 2004 77 1053 1056 15569651 10.1259/bjr/65044256
Borgognoni L Urso C Vaggelli L Brandani P Gerlini G Sentinel node detection in melanoma patients using computer-assisted gamma probe with adjustable collimation and specifically designed forceps Oral presentation, International Sentinel Node Congress, Los Angeles, CA, December 3–6, 2004
Pelosi E Bello M Giors M Ala A Giani R Bussone R Bisi G Sentinel lymph node detection in patients with early-stage breast cancer: comparison of periareolar and subdermal/peritumoral injection techniques J Nucl Med 2004 45 220 225 14960639
Kern KA Lymphoscintigraphic anatomy of sentinel lymphatic channels after subareolar injection of Technetium 99m sulfur colloid J Am Coll Surg 2001 193 601 608 11768675 10.1016/S1072-7515(01)01068-7
Uren RF Thompson JF Howman-Giles R Sentinel nodes Interval nodes, lymphatic lakes, and accurate sentinel node identification Clin Nucl Med 2000 25 234 236 10698433 10.1097/00003072-200003000-00025
Tanis PJ van Sandick JW Nieweg OE Valdes Olmos RA Rutgers EJ Hoefnagel CA Kroon BB The hidden sentinel node in breast cancer Eur J Nucl Med Mol Imaging 2002 29 305 311 12002703 10.1007/s00259-001-0732-y
Krynyckyi BR Chun H Kim HH Eskandar Y Kim CK Machac J Factors affecting visualization rates of internal mammary sentinel nodes during lymphoscintigraphy J Nucl Med 2003 44 1387 1393 12960181
Estourgie SH Nieweg OE Olmos RA Rutgers EJ Kroon BB Lymphatic drainage patterns from the breast Ann Surg 2004 239 232 237 14745331 10.1097/01.sla.0000109156.26378.90
Krynyckyi BR Shim J Kim CK Internal mammary chain drainage of breast cancer [letter] Ann Surg 2004 240 557 author reply 558. 15319727 10.1097/01.sla.0000138728.70176.7f
Feldman SM Krag DN McNally RK Moor BB Weaver DL Klein P Limitation in gamma probe localization of the sentinel node in breast cancer patients with large excisional biopsy J Am Coll Surg 1999 188 248 254 10065813 10.1016/S1072-7515(98)00306-8
Keshtgar MR Chicken DW Waddington WA Raven W Ell PJ A training simulator for sentinel node biopsy in breast cancer: a new standard Eur J Surg Oncol 2005 31 134 40 15698728 10.1016/j.ejso.2004.08.008
Krynyckyi BR Singh G Colon D Kim CK Travis A Kim SC Machac J Letter to the editor [letter] Eur J Surg Oncol 2005 31 805 806 15975758 10.1016/j.ejso.2005.04.015
Gentilini O Cremonesi M Trifiro G Ferrari M Baio SM Caracciolo M Rossi A Smeets A Galimberti V Luini A Tosi G Paganelli G Safety of sentinel node biopsy in pregnant patients with breast cancer Ann Oncol 2004 15 1348 1351 15319240 10.1093/annonc/mdh355
Keleher A Wendt R 3rdDelpassand E Stachowiak AM Kuerer HM The safety of lymphatic mapping in pregnant breast cancer patients using Tc-99m sulfur colloid Breast J 2004 10 492 495 15569204 10.1111/j.1075-122X.2004.21503.x
Gennari R Rotmensz N Perego E Santos GD Veronesi U Sentinel node biopsy in elderly breast cancer patients Surg Oncol 2004 13 193 196 15615656 10.1016/j.suronc.2004.08.011
Kennedy RJ Kollias J Gill PG Bochner M Coventry BJ Farshid G Removal of two sentinel nodes accurately stages the axilla in breast cancer Br J Surg 2003 90 1349 1353 14598413 10.1002/bjs.4298
Schrenk P Rehberger W Shamiyeh A Wayand W Sentinel node biopsy for breast cancer: does the number of sentinel nodes removed have an impact on the accuracy of finding a positive node? J Surg Oncol 2002 80 130 136 12115795 10.1002/jso.10112
McMasters KM Wong SL Chao C Comment on the article "Highest isotope count does not predict sentinel node positivity in all breast cancer patients," by Martin et al., August 2001, Annals of Surgical Oncology [letter] Ann Surg Oncol 2002 9 317 11923141 10.1245/aso.2002.9.3.317
Camp ER Cendan JC Feezor R Lind DS Wilkinson E Copeland EM The hottest sentinel lymph node is not always the positive node Am Surg 2004 70 475 478 15212397
Uren RF Howman-Giles R Chung D Thompson JF Nuclear medicine aspects of melanoma and breast lymphatic mapping Semin Oncol 2004 31 338 348 15190491 10.1053/j.seminoncol.2004.03.007
Haigh PI Hansen NM Giuliano AE Edwards GK Ye W Glass EC Factors affecting sentinel node localization during preoperative breast lymphoscintigraphy J Nucl Med 2000 41 1682 1688 11037998
Uren RF Howman-Giles RB Chung D Thompson JF Role of lymphoscintigraphy for selective sentinel lymphadenectomy Cancer Treat Res 2005 127 15 38 16209076
Thompson JF Uren RF Scolyer RA Stretch JR Selective sentinel lymphadenectomy: progress to date and prospects for the future Cancer Treat Res 2005 127 269 287 16209088
Krynyckyi BR Zhang ZY Kim CK Lipszyc H Mosci K Machac J Effect of high specific-activity sulfur colloid preparations on sentinel node count rates Clin Nucl Med 2002 27 92 95 11786736 10.1097/00003072-200202000-00003
Diaz NM Vrcel V Centeno BA Muro-Cacho C Modes of benign mechanical transport of breast epithelial cells to axillary lymph nodes Adv Anat Pathol 2005 12 7 9 15614159 10.1097/01.pap.0000151267.34438.a1
Hampton T Surgeons "vote with their feet" for sentinel node biopsy for breast cancer staging JAMA 2003 290 3053 3054 14679253 10.1001/jama.290.23.3053
|
16277655
|
PMC1308847
|
CC BY
|
2021-01-04 16:38:36
|
no
|
Int Semin Surg Oncol. 2005 Nov 8; 2:25
|
utf-8
|
Int Semin Surg Oncol
| 2,005 |
10.1186/1477-7800-2-25
|
oa_comm
|
==== Front
Int Semin Surg OncolInternational seminars in surgical oncology : ISSO1477-7800BioMed Central London 1477-7800-2-261631367410.1186/1477-7800-2-26Case ReportSkin and fat necrosis of the breast following methylene blue dye injection for sentinel node biopsy in a patient with breast cancer Salhab M [email protected] sarakbi W [email protected] K [email protected] St Georges and The Princess Grace Hospitals, London, UK2005 28 11 2005 2 26 26 27 9 2005 28 11 2005 Copyright © 2005 Salhab et al; licensee BioMed Central Ltd.2005Salhab et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Sentinel lymph node biopsy (SLNB) is a simple technique that uses subdermal or peri-tumoral injection of vital blue dye and/or radioactive isotope to identify the first lymph node(s) draining the primary tumor. It has been shown to accurately predict axillary node status in patients with clinically node negative breast cancer. The SLNB is emerging as a new standard of care in patients with early breast cancer. However, the use of methylene blue (MB) dye can be associated with a number of local complications due to its tissue reactive properties. We report a rare case of skin and fat necrosis followed by a dry gangrene of the skin in a female patient with breast cancer who underwent SLNB localization using peri-tumoral injection of MB dye in another institution. This case and literature review suggest that the use of MB dye for SLNB identification should be avoided and replaced with alternative types of blue dye such as Patent Blue V preferably in conjunction with a radioactive isotope tracer.
Sentinel lymph nodemethylene blue dyebreast cancerskin necrosis
==== Body
Introduction
Sentinel lymph node biopsy (SLNB) is emerging as a new standard in the treatment of patients with operable breast cancer. It is a simple technique that uses subdermal or peri-tumoral injection of vital blue dye and/or radioactive isotope to identify the first lymph node(s) draining the primary tumor. SLNB is a reliable and minimally invasive procedure, which accurately predicts axillary node status in patients with clinically node negative breast cancer [1]. Localization of the sentinel lymph node using the intradermal, subareolar or peritumoural injection of a vital blue dye is widely practiced [2]. However, local and systemic complications secondary to the use of the dye have been reported.
We report a rare case of severe skin and fat necrosis secondary to the injection of methylene blue dye in a patient with T1 breast cancer. This complication occurred in a different institution.
Case report
A postmenopausal woman was referred to the senior author (KM) following breast conserving surgery for 17 mm invasive ductal carcinoma of the right breast. Localization of the sentinel lymph node was performed using the dual localization technique. Intradermal injection of technetium-labled sulphur colloid was performed over the tumour site. The following day, the patient had the MB dye injected in the peritumoral area. The operation was uneventful.
Three days post operatively and upon removal of the wound dressing, the lateral aspect of the breast skin exhibited a rectangular erythematous violaceous surface which developed into a dry gangrenous area a few days later (Figure 1). We believe that the patient developed skin and fat necrosis secondary to the MB dye injection. This complication may have been caused by a localised tissue reaction initiated by, or involving the dye.
Figure 1 Skin and fat necrosis of the right breast secondary to injection of methylene blue dye for SLNB.
Discussion
The technique of blue dye mapping was first described for breast cancer by Giuliano et al [3]. Isosulfan blue dye has been traditionally used the dye used for SLNB for breast cancer. However, its use was associated with a significant number of allergic reactions [4], some of which are life threatening. Because methylene blue dye has been shown to be equally effective and does not pose a serious risk of severe allergic and hypersensitivity reactions, it was regarded as an acceptable substitute for isosulfan blue dye for SLNB [5-8]. Although, the use of the MB dye for SLNB in breast cancer has fewer allergic reactions, its use has been associated with a number of local and systematic complications. Stradling et al, was the first to report adverse skin reactions to methylene blue dye in patients with breast cancer [9]. In addition, skin eruptions and rashes [10], subcutaneous tissue necrosis and abscess formation [11] have been reported in association with the injection of this dye. Furthermore, capsular contraction following breast reconstruction using an implant with intense blue discoloration of the prosthesis was reported in a patient in whom methylene blue dye was used to identify the sentinel lymph node [12].
In our reported case, severe skin and fat necrosis complicated the peri-tumoral injection of methylene blue dye; This might be due to that methylene blue dye may induce an early foreign body-type reaction characterized by ischemic ulceration, fibrinoid necrosis with eosinophilic infiltration [13].
Therefore, we recommend the use of Patent Blue V dye instead of MB for SLNB localization in patients with breast cancer in order to avoid such significant complications which may delay subsequent treatment. Patent Blue dye has been reported to cause minor local complications in form of long-term discoloring of the skin at the site of injection [14]. Although no cases of severe local tissue necrosis has been reported in association with Patent Blue V dye, however, anaphylactic shock has been observed following its injection for SLNB localization [15,16]. The risk of allergic reactions can be reduced by using corticosteroids and antihistamines [4,17,18]
In conclusion, the use of MB dye for SLNB identification should be avoided and replace with alternative types of blue dye such as Patent Blue V preferably in conjunction with a radioactive isotope tracer.
==== Refs
Singh-Ranger G Mokbel K The sentinel node biopsy is a new standard of care for patients with early breast cancer Int J Fertil Womens Med 2004 49 225 7 15633480
Mokbel K Mostafa A The role of subareolar blue dye in identifying the sentinel node in patients with invasive breast cancer Curr Med Res Opin 2001 17 93 5 11759188 10.1185/030079901317010720
Giuliano AE Kirgan DM Guenther JM Morton DL Lymphatic mapping and sentinel lymphadenectomy for breast cancer Ann Surg 1994 220 391 401 8092905
Cimmino VM Brown AC Szocik JF Pass HA Moline S De SK Domino EF Allergic reactions to isosulfan blue during sentinel node biopsy – a common event Surgery 2001 130 439 42 11562667 10.1067/msy.2001.116407
Simmons RM Smith SM Osborne MP Methylene blue dye as an alternative to isosulfan blue dye for sentinel lymph node localization Breast J 2001 7 181 3 11469932 10.1046/j.1524-4741.2001.007003181.x
Thevarajah S Huston TL Simmons RM A comparison of the adverse reactions associated with isosulfan blue versus methylene blue dye in sentinel lymph node biopsy for breast cancer Am J Surg 2005 189 236 9 15720998 10.1016/j.amjsurg.2004.06.042
Tuttle TM Technical advances in sentinel lymph node biopsy for breast cancer Am Surg 2004 70 407 13 15156948
Eldrageely K Vargas MP Khalkhali I Venegas R Burla M Gonzalez KD Vargas HI Sentinel lymph node mapping of breast cancer: a case-control study of methylene blue tracer compared to isosulfan blue Am Surg 2004 70 872 5 15529840
Stradling B Aranha G Gabram S Adverse skin lesions after methylene blue injections for sentinel lymph node localization Am J Surg 2002 184 350 2 12383900 10.1016/S0002-9610(02)00945-5
Raimer SS Quevedo EM Johnston RV Dye rashes Cutis 1999 63 103 106 10071743
Borgstein PJ Meijer S Pijpers R Intradermal blue dye to identify the sentinel lymph node in breast cancer Lancet 1997 349 1668 1669 9186389 10.1016/S0140-6736(05)62634-7
Singh-Ranger G Mokbel K Capsular contraction following immediate reconstructive surgery for breast cancer – An association with methylene blue dye Int Semin Surg Oncol 2004 11, 1 3 15285809 10.1186/1477-7800-1-3
Lane KL Vallera R Washington K Gottfried MR Endoscopic tattoo agents in the colon. Tissue responses and clinical implications Am J Surg Pathol 1996 20 1266 70 8827034 10.1097/00000478-199610000-00013
Govaert GA Oostenbroek RJ Plaisier PW Prolonged skin staining after intradermal use of patent blue in sentinel lymph node biopsy for breast cancer Eur J Surg Oncol 2005 31 373 5 15837042 10.1016/j.ejso.2004.12.009
Woltsche-Kahr I Komericki P Kranke B Brabek E Horn M Schuller-Petrovic S Richtig E Aberer W Anaphylactic shock following peritumoral injection of patent blue in sentinel lymph node biopsy procedure. 2 Eur J Surg Oncol 2000 26 313 4 10753539 10.1053/ejso.1999.0888
Mostafa A Carpenter R Anaphylaxis to patent blue dye during sentinel lymph node biopsy for breast cance Eur J Surg Oncol 2001 27 610 11520102 10.1053/ejso.2001.1174
Aubard Y Mollard J Ducloux T Monteil J Fermeaux V Desfougeres M Gana J Serdouma E Detection of the sentinel lymph node under local anaesthesia in early-stage breast cancer: feasibility study in a series of 78 unselected patients Eur J Gynaecol Oncol 2004 25 178 82 15032276
Dubost JL Chevallier H Allergic reactions to patent blue violet: mechanisms, frequency and treatment Phlebologie 35 739 46 1982 Jul-Sep 6182572
|
16313674
|
PMC1308848
|
CC BY
|
2021-01-04 16:38:37
|
no
|
Int Semin Surg Oncol. 2005 Nov 28; 2:26
|
utf-8
|
Int Semin Surg Oncol
| 2,005 |
10.1186/1477-7800-2-26
|
oa_comm
|
==== Front
J Autoimmune DisJournal of Autoimmune Diseases1740-2557BioMed Central London 1740-2557-2-111630305410.1186/1740-2557-2-11HypothesisPhenytoin as a novel anti-vitiligo weapon Namazi MR [email protected] Dermatology Department, Shiraz University of Medical Sciences, Shiraz, Iran2005 22 11 2005 2 11 11 12 4 2005 22 11 2005 Copyright © 2005 Namazi; licensee BioMed Central Ltd.2005Namazi; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Vitiligo is a psychologically devastating clinical conundrum which affects approximately 1% of the general population. The exact cause of the illness is an enigma, but several hypotheses about its pathogenesis are advanced.
The autoimmune hypothesis proposes an autoimmune attack against melanocytes. Although anti-melanocyte autoantibodies have been demonstrated in vitiligo, recent research casts doubt on their pathogenic role and instead supports the involvement of cell-mediated autoimmune response in the pathobiology of this disorder, which is characterized by increase of suppressor T-cells and decrease of the helper/suppressor ratio in association with the presence of type-1 cytokine secreting cytotoxic T cells in the vicinity of disappearing melanocytes.
The neural hypothesis proposes that increased release of norepinephrine, a melanocytotoxin, from the autonomic nerve endings in the microenvironment of melanocytes injures these cells. Moreover, norepinephrine induces the catecholamine degrading enzyme monoamine oxidase (MAO), which favors the formation of toxic levels of hydrogen peroxide in the vicinity of melanocytes.
Another theory suggests that abnormal permeability of melanosome membrane, which normally prevents the diffusion of toxic melanin precursors into the cytoplasm, may cause melanocyte damage.
Phenytoin, the widely-used anticonvulsant, has been employed both topically and systemically in the treatment of some dermatological disorders. The drug has been shown to significantly suppress mitogen-induced activation of lymphocytes and cytotoxic T lymphocyte activity and to polarize the immune response toward the type-2 pathway. It also significantly decreases suppressor T cells and increases the helper/suppressor ratio.
At high concentrations, the drug inhibits the release of norepinephrine and the activity of MAO. Moreover, phenytoin is suggested to interact with membrane lipids, which may promote stabilization of the membranes.
The hydantoin moiety of phenytoin exerts a direct stimulatory action on melanocytes; facial hyperpigmentation is a recognized side effect of orally administered phenytoin.
Altogether, the above evidence suggests that phenytoin could be therapeutically effective against vitiligo. As phenytoin stimulates collagen production and inhibits its breakdown, its concomitant use with topical steroids could prevent steroid-induced skin atrophy while potentiating the anti-vitiligo effect of these agents.
DilantinTreatmentMelanocyteImmunityCytokines
==== Body
Vitiligo, a psychologically devastating and frequently recalcitrant skin disorder, affects approximately 1% of the general population [1]. Theories concerning the cause of vitiligo have focused on three different mechanisms: auto-immune, neural, and autocytotoxic. The autoimmune hypothesis proposes an autoimmune attack against melanocytes. The autocytotoxic theory postulates that cytotoxic precursors to melanin synthesis accumulate in melanocytes, causing cell death. The neural hypothesis proposes that the elevated levels of some neurotransmitters and catecholamine degrading enzymes injure melanocytes [2].
The important aspects of each theory pertinent to the present paper are discussed below.
I) Evidence for involvement of cell-mediated immunity in the pathogenesis of vitiligo
Although melanocyte-specific autoantibodies have been demonstrated in vitiligo, their pathogenic role remains uncertain [1]. Though these autoantibodies have been shown to be able to damage pigment cells both in vitro and in vivo, more recently it has been reported that specific autoantibodies for tyrosinase are present at low frequency in the sera of vitiligo patients, and that such autoantibody titers may be found both in vitiligo-like depigmentation (melanoma-associated depigmentation) and in healthy controls [3]. The patchy distribution of cutaneous depigmentation and the most frequent symmetrical distribution of lesions corroborate the concept that autoimmune melanocyte damage is caused by clones of lymphocytes with affinities for specific areas of skin [1]. In fact, high frequencies of circulating melanocyte-specific CD8+ T cells expressing skin-homing capacity and exerting anti-melanocyte cytotoxicity in vitro have been found in vitiligo patients [3]. Moreover infiltrating activated T lymphocytes have been observed at the periphery of vitiligo lesions and a recent immunopathologic study of lesional skin of vitiligo patients noted a high frequency of cutaneous lymphocyte antigen-positive activated cytotoxic T cells clustered in perilesional skin adjacent to the disappearing melanocytes [4]. An increased level of soluble interleukin (IL)-2 receptor, IL-6 and IL-8 has been observed in vitiligo patients, which further suggests that T cell activation may be a component in vitiligo pathogenesis [5].
Notably, perilesional T-cell clones exhibited a predominant type-1-like cytokine secretion profile, whereas the degree of type-1 deviation in uninvolved skin-derived T-cell clones correlated with the process of microscopically observed melanocyte destruction in situ. Detailed analysis of broad spectrum of cytokines produced by lesional- and nonlesional-derived CD4+ and CD8+ T-cell clones confirmed deviation toward type-1-like in both CD4 and CD8 compartments, which mirrored depigmentation process observed locally in the skin [4]. The facts that vitiligo is found more frequently in patients with metastatic melanoma and is associated with an improved prognosis, and that vitiligo-like depigmentation has been observed following immunotherapy of melanoma, further support a crucial role for cell-mediated immunity in the pathogenesis of the disorder [5].
It deserves noting that the peripheral blood of patients with vitiligo shows a significant decrease in helper cells and helper/suppressor ratios in comparison with control subjects [6].
An association between CD8+ T lymphocyte reactivity to the melanocyte antigen gp100, and to a lesser extent Melan A/MART-1, and vitiligo has been demonstrated. Furthermore, the disease activity appears correlated with reactivity to gp100 [5].
II) Involvement of increased norepinephrine (NE) and monoamine oxidase (MAO) levels in the pathogenesis of vitiligo
It has recently been demonstrated that axon terminals and epidermal melanocytes make contact via chemical synapses in human skin [2,7].
An increased release of catecholamine from the autonomic nerve endings in the microenvironment of melanocytes in the vitiliginous areas has been reported. NE is known for having direct and indirect toxic effects on melanocytes [2,7].
Direct actions include interacting with cellular sulfhydryl groups, enzyme inhibition, impairing mitochondrial calcium uptake, and forming some cytotoxic products such as free radicals. Indirect actions include activating α-receptors of skin arterioles causing a severe vasoconstriction, thereby producing toxic oxygen radicals due to hypoxia [2,7]. Moreover, elevated level of NE secreted by keratinocytes or by nerve endings induces the catecholamine degrading enzyme monoamine oxidase (MAO). Increased MAO activity favors the formation of toxic levels of hydrogen peroxide, which is not buffered due to the low catalase activity in vitiliginous skin [2,7].
III) Abnormal permeability of melanosome membrane as a potential pathogenic factor in vitiligo
Tyrosine, a derivative of phenol, is oxidized into melanin via a complex series of oxidative reactions, often accompanied by electronic rearrangements within molecules. Some of these products are unstable and capable of forming radicals that react with other molecules in the cell. It is thought that melanin synthesis is confined within the melanosome to prevent these melanin precursors from diffusing into the cell where they might disrupt essential metabolic pathways, leading to the destruction of the pigment cell.
Defects either inherited or acquired in the membranes of melanosomes resulting in the demise of melanocytes are supposed to cause vitiligo [8].
IV) Phenytoin exerts an inhibitory effect on the pathomechanism behind vitiligo but a direct stimulatory action on melanocytes
Phenytoin (dilantin), a diphenyl-substituted hydantoin, is a widely prescribed anticonvulsant agent which has been used in cutaneous medicine in the treatment of ulcers, inflammatory conditions, and epidermolysis bullosa [9]. Interestingly, this agent possesses a myriad of effects which make it potentially effective against vitiligo:
Phenytoin is documented to significantly suppress mitogen-induced activation of lymphocytes [10,11], cytotoxic T lymphocyte activity [11-14] and cell-mediated immunity [15], polarizing the immune response toward a type-2-like cytokine secretion profile [14]. Additionally, it significantly decreases suppressor T-cells and increases the helper/suppressor ratio [16,17].
Phenytoin inhibits the production of superoxide anion by immune cells [18] and significantly improves rheumatoid arthritis, a cell-mediated autoimmune disorder [19].
Interestingly, some in vitro studies have shown that this drug, at high concentrations, which may also be reached by the topical application, inhibits the release of NE and inhibits the activity of MAO [20]. it is also speculated that the agent may interact with membrane lipids (and conceivably with melanosome membrane lipids as well), which could promote the stabilization of the membranes [20].
Melasma (chloasma), the facial hypermelanosis produced by hyperfunctional melanocytes, is a recognized side effect of orally administered phenytoin [9,21]. It has been suggested that the hydantoin moiety exerts a direct action on the melanocytes; it has been shown to expand melanophores in amphibian larvae [21].
In conclusion, given its inhibitory effect on cell-mediated immunity, NE release, and MAO activity and also in view of its potential ability to stabilize melanosome membrane and to stimulate melanocytes, phenytoin may prove to be effective against vitiligo. Importantly, as phenytoin facilitates collagen deposition and inhibits collagenase activity [9], its concomitant topical use with steroids may prevent steroid-induced skin atrophy while potentiating the anti-vitiligo effect of these agents. The latter effect could occur through phenytoin's ability to directly stimulate melanocyte proliferation, which is lacked by steroids. Moreover, the immunomodulatory effect of this agent, which may be employed topically in relatively high concentrations without producing the untoward systemic effects, may permit the reduction of the concentration of the topical steroids (steroid-sparing effect) and thus their frequently-occurring dermatologic side effects. As the closing comment, topically-administered phenytoin was shown to exert significant immunomodulatory activity and be effective against lichen planus [22], which could reasonably be the case in vitiligo as well.
V) Testing this hypothesis
This hypothesis can be tested using animal models of vitiligo such as the Smyth chicken. The Smyth chicken expresses the major features of human vitiligo – the delayed development of cutaneous amelanosis, most commonly appearing during adolescence or at young adulthood, and variable in occurrence, location and extent [23]. The etiology of vitiligo in the Smyth chicken appears to include an inherent defect in the melanocytes and an associated autoimmune response [24]. Though the Smyth chicken does not incorporate all the genetic and environmental factors which may do or cause vitiligo in humans, it is an excellent model for the study of human vitiligo in that it parallels the human condition in variability, complexity, mode of expression, and polyfactorial etiology. As complement to animal studies, this hypothesis could be tested most rigorously in patients suffering from bilateral lesions by applying a topical preparation of phenytoin to the lesion(s) of one side of the body and a topical placebo to the contralateral lesion(s), then comparing the results (double-blind, placebo-controlled, bilateral paired comparison method).
==== Refs
Mantovani S Garbelli S Palermo B Campanelli R Brazzelli V Borroni G Molecular and functional bases of self-antigen recognition in long-term persistent melanocyte-specific CD8+ T cells in one vitiligo patient J Invest Dermatol 2003 121 308 14 12880423 10.1046/j.1523-1747.2003.12368.x
Namazi MR Prescribing cyclic antidepressants for vitiligo patients: which agents are superior, which are not? Psychother Psychosom
Palermo B Campanelli R Garbelli S Mantovani S Lantelme E Brazzelli V Specific cytotoxic T lymphocyte response against Melan-A/MART 1, tyrosinase and GP100 in vitiligo by the use of major histocompatibility complex/peptide tetramers: the role of cellular immunity in the etiopathogenesis of vitiligo J Invest Dermatol 2001 117 326 32 11511311 10.1046/j.1523-1747.2001.01408.x
Wankowicz-Kalinska A van den Wijngaard RM Tiggers BJ Westerhof W Ogg GS Cerundolo V Immunopolarization of CD4+ and CD8+ T cells to type-1-like is associated with melanocyte loss in human vitiligo Lab Invest 2003 83 683 95 12746478
Mandelcorn-Monson RL Shear NH Yau E Sambhara S Barber BH Spaner D Cytotoxic T lymphocyte reactivity to gp100, melan A/MART 1, and tyrosinase, in HLA-A2-positive vitiligo patients J Invest Dermatol 2003 121 550 6 12925214 10.1046/j.1523-1747.2003.12413.x
Grimes PE Ghoneum M Stockton T Payne C Kelly AP Alfred L T cell profiles in vitiligo J Am Acad Dermatol 1986 14 196 201 2936773
Orecchia GE Hann SK, Nordlund JJ Neural pathogenesis Vitiligo 2000 1 Oxford: Blackwell Scientific Publications 142 50
Han SK Chun WH Hann SK, Nordlund JJ Autocytotoxic hypothesis for the destruction of melanocytes as the cause of vitiligo Vitiligo 2000 1 Oxford: Blackwell Scientific Publications 137 41
Scheinfeld N Phenytoin in cutaneous medicine: Its uses, mechanisms, and side effects Dermatol Online J 2003 9 6 12952753
Kikuchi K McCormick CI Neuwelt EA Immunosuppression by phenytoin: implication for altered immune competence in brain-tumor patients J Neurosurg 1984 61 1085 90 6502237
Okamoto Y Shimizu K Tamura K Yamada K Matsui Y Hayakawa T [Effects of anticonvulsants on cellular immunity] No To Shinkei 1989 41 299 304 2503016
Okamoto Y Shimizu K Tamura Y Miyao Y Yamada M Tsuda N Effects of phenytoin on cell-mediated immunity Cancer Immunol Immunother 1988 26 176 9 3258793 10.1007/BF00205612
Okamoto Y Shimizu K Tamura Y Miyao Y Yamada M Matsui Y [Effects of phenytoin on cell-mediated immunity] No To Shinkei 1987 39 931 6 3501727
Okada K Sugiura T Kuroda E Tsuji S Yamashita U Phenytoin promotes Th2 type immune response in mice Clin Exp Immunol 2001 124 406 13 11472401 10.1046/j.1365-2249.2001.01491.x
Chiu HC Hsieh KH Hung TP Young MC [Humoral and cell-mediated immunities in epileptic patients] Zhonghua Min Guo Wei Sheng Wu Ji Mian Yi Xue Zazhi 1982 15 30 7
Basaran N Kansu E Hincal F Serum immuno-globulins, complement levels and lymphocyte subpopulations in phenytoin-treated epileptic patients Immunopharmacol Immunotoxical 1989 11 335 46
Basaran N Hincal F Kansu E Ciger A Humoral and cellular immune parameters in untreated and phenytoin- or carbamazepine-treated epileptic patients Int J Immunopharmacol 1994 16 1071 7 7705969 10.1016/0192-0561(94)90087-6
Marcoli M Ricevuti G Fasani F Mazzone A Baiguera R Tarabini L Assessment of lymphocyte and phagocytic cell function in healthy volunteers undergoing short-term phenytoin administration Int J Immunopharmacol 1987 9 903 12 3429077 10.1016/0192-0561(87)90006-3
Grindulis KA Nichol FE Oldham R Phenytoin in rheumatoid arthritis J Rheumatol 1986 13 1035 9 3560089
Porter RJ Meldrum BS Katzung BG Antiepileptic drugs Basic and Clinical Pharamacology 1992 5 Connecticut: Appleton & Lange 333 6
Bleehen SS Champion RH, Burton JL, Burns DA, Breathnach SM Disorders of skin color Rook/Wilkinson/Ebling Textbook of Dermatology 1998 6 London: Blackwell Scientific Publications 1785
Bogaert H Sanchez E Lichen planus: Treatment of thirty cases with systemic and topical phenytoin Int J Dermatol 1990 29 157 8 2323879
Lamoreux L Boissy RE Hann SK, Nordlund JJ Animal models Vitiligo 2000 1 Oxford: Blackwell Scientific Publications 281 297
Smyth JR Jr The Smyth chicken: model of autoimmune amelanosis Critical Reviews in Poultry Science 1989 1 9
|
16303054
|
PMC1308849
|
CC BY
|
2021-01-04 16:38:03
|
no
|
J Autoimmune Dis. 2005 Nov 22; 2:11
|
utf-8
|
J Autoimmune Dis
| 2,005 |
10.1186/1740-2557-2-11
|
oa_comm
|
==== Front
J Autoimmune DisJournal of Autoimmune Diseases1740-2557BioMed Central London 1740-2557-2-91628008610.1186/1740-2557-2-9ResearchCo-occurrence of autoimmune thyroid disease in a multiple sclerosis cohort Sloka JS [email protected] Pryse-WEM [email protected] M [email protected] C [email protected] Department of Neurology, Memorial University of Newfoundland, St John's, Canada2 Department of Endocrinology, Memorial University of Newfoundland, St John's, Canada2005 9 11 2005 2 9 9 6 9 2005 9 11 2005 Copyright © 2005 Sloka et al; licensee BioMed Central Ltd.2005Sloka et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Multiple sclerosis (MS), Hashimoto's disease and Graves' disease are autoimmune diseases that may share similar pathogenic mechanisms. The co-occurrence rates and demographic characteristics of Graves' disease and Hashimoto's disease (HT) in our MS population are compared with the general population.
Methods
The prevalence of thyroid disease in our MS patients was determined by chart review and survey. Previous diagnosis of thyroid disease, age at diagnosis, treatment used, and about the use of disease modifying medications used to treat their MS were asked. Chart reviews were used to estimate the population prevalence of Graves' disease and Hashimoto's disease and to estimate the demographics of patients with thyroid disease.
Results
A significant co-occurrence of Graves' disease with MS (p = 0.002), and a non-significant co-occurrence of Hashimoto's disease were noted (p = 0.097). No difference in the age of onset or gender of thyroid disease in MS patients compared to the general population was found.
Conclusion
There is a significant co-occurrence in patients with MS and Graves' disease, and a trend to co-occurrence in patients with MS and Hashimoto's disease. There are no differences in the demographics of patients with thyroid disease in our MS patients compared to the general population.
==== Body
Introduction
Autoimmune diseases, sometimes defined as a clinical syndrome caused by the activation of T cells or B cells, or both, in the absence of an ongoing infection or other discernible cause [1], comprise a heterogeneous group of disorders wherein alterations in the immune system may result in a spectrum of disease that either targets specific organs or affects the body systemically. Even though each purported autoimmune disease affects disparate organs and systems, the existence of fundamental pathophysiological mechanisms is hypothesized. Recent such epidemiological studies have shown an increased statistical susceptibility for people with one autoimmune disease to develop another [2-5]. This provocative increased relative risk of acquiring a second autoimmune disease may be due to a genetic susceptibility that affects both diseases, an environmental trigger that initiates both diseases, the alteration of the body's homeostasis by one disease that creates susceptibility to another, or some as yet undefined shared mechanism.
Some studies have shown that autoimmune diseases "cluster together"[5]. Specifically, several studies have shown an increased co-occurrence of MS with Hashimoto's thyroiditis (HT) as compared to the general population [3,4,6] as well as an increased co-occurrence of MS with Graves' disease [7] while other have not [2]. No comment on the age of onset of thyroid disease in MS patients compared with the general population has been previously made. Some studies have shown that the risk of acquiring autoimmune thyroid disease has been shown to increase after treatment with some MS disease modifying agents [8]. The population co-occurrence of MS and autoimmune thyroid disease has not been studied after accounting for previous disease modifying therapy[5].
We have undertaken an epidemiological study of MS and autoimmune thyroid diseases in Newfoundland and Labrador (NL) to elucidate some of the epidemiological features of any co-occurrence found amongst these diseases. Presumably, if one is susceptible to more than one disease, then features such as clinical onset of disease may be seen earlier in those more susceptible, especially if one disease provokes another.
NL, with its centralized and comprehensive medical records system and its unique population [9,10] is an excellent place to perform such epidemiological studies. The prevalence of MS was previously determined [11] with near complete case ascertainment. However, the prevalence of thyroid disease in the general population of NL was unknown. Therefore, an estimate of the prevalence of thyroid disease in NL was also undertaken in order to compare the prevalence of thyroid disease in the MS population with the prevalence of thyroid disease in the general NL population and to compare their demographics.
Methods
An incidence and prevalence study for the Canadian province of Newfoundland and Labrador has recently been completed (Dec 31, 2001), demonstrating an MS population prevalence for MS of 94.4 per 100,000 [11]. A database was created with patient information for 493 confirmed living cases [11]. Complete chart reviews were conducted on each confirmed case to determine if these patients had concomitant thyroid disease. Careful chart reviews were performed in the patients' neurology charts and their inpatient hospital charts. As well, electronic searches through the hospital information systems for the four main community hospitals were conducted for both case ascertainment and diagnosis confirmation of MS and thyroid disease. MS diagnosis type, thyroid diagnosis type, and age at onset of disease were captured.
In October 2003, a survey was mailed to the living diagnosed patients from the prevalence study whose current addresses could be confirmed, with a subsequent mailer to those that did not initially respond after 6 weeks. The survey asked questions relating to concomitant thyroid disease and age of initial symptoms (Table 1). This survey and all case searches were conducted with the approval of the Human Investigations Committee of Memorial University.
Table 1 Survey questions related to thyroid disease
Have you ever been treated for thyroid disease in the past?
Yes No
If yes (multiple yes answers are possible)
Was this called Graves' disease? No Yes
Was this called Hashimoto's disease? No Yes
Was this called hyperthyroidism? No Yes
Was this called hypothyroidism? No Yes
Did you ever take Synthroid? No Yes
Did you ever take Eltroxin? No Yes
Did you ever take propylthiouricil? No Yes
Did you ever take methimazole? No Yes
Did you ever have radiation treatment for your thyroid? No Yes
At what age were you diagnosed with any of the above? ________ years old
To determine the prevalence of thyroid disease in the general population, a representative population sample was sought whereby complete case histories were taken on clinic patients and bias towards a known predisposition to thyroid disease was minimized. The general neurology clinic was used for case ascertainment of thyroid disease in a representative general population after it was determined by thorough literature search that there is no known association of thyroid disease in patients with primary headache disorders or injuries (eg falls, motor vehicle accidents (MVA), sports injuries). Therefore, a retrospective chart review was conducted through all the patient charts that have been evaluated in the general neurology clinic of two of the authors (WPP and MS) over the previous 8 years. Headache type/injury source, gender, age at clinic visit and concomitant thyroid disease were captured.
In order to minimize case ascertainment discrepancies between the MS population sample and the general population sample, a followup electronic chart review was performed on both the general and MS population cases. As per the previous MS incidence and prevalence study, complete physician billing records towards all thyroid disease and MS for the entire province were requested from the provincial medical insurance plan via an official government Order in Council. This aided in the confirmation of thyroid disease in both sample populations and minimized the number of missed thyroid histories in the original chart review. These billing records also increased accuracy of case ascertainment and thus reduced referral and ascertainment bias. Confirmation of thyroid disease in both the MS population and the general population sample was completed for all diagnoses via review of patient charts and electronic records searching specifically for thyroid stimulating hormone (TSH) receptor autoantibody tests, nuclear medicine studies, and diagnosis by an endocrinologist or internist. Patients with MS were excluded from the general population database and patients from St Pierre et Miquelon, a neighbouring territory, were excluded from databases due to inconsistent referral.
The diagnostic criteria used by endocrinologists for Graves' disease has evolved over time[12]. All patients with diagnosed Graves' disease in our study had increased free thyroxine and decreased TSH. All patients were evaluated using either 99 mTc radionucleotide scintigraphy or 131I uptake and scan. The diagnosis in some patients was supported by positive TSH receptor autoantibodies. Patients with Graves' disease were usually treated with anti-thyroid drugs, and subsequent radioablation therapy if required. The diagnosis of Hashimoto's disease was based on low or low normal free T4 with elevated TSH, and was supported by the presence of elevated anti-thyroperoxidase autoantibodies, and anti-thyroglobulin autoantibodies. In addition, on clinical examination, a firm goiter in the context of the aforementioned laboratory findings corroborated the diagnosis of Hashimoto's disease. These patients were generally treated with thyroid hormone replacement therapy.
Through initial sampling of patient charts, it was estimated that a total of 500 patients would be found. For a estimated prevalence of Hashimoto's thyroiditis of 2% [13] (yielding 10 of 250 patients) and Graves' disease of 0.8% [13,14] (yielding 4 of 250 patients), it was decided that this would not yield a large enough sample size to give a clear indication of the demographics of thyroid disease in NL. Therefore, another retrospective chart review was conducted in order to estimate the demographics of Graves' and Hashimoto's disease. All the endocrinology clinic patient charts of another author (CJ) were reviewed from the previous 15 years to determine the demographics of patients in the province with (autoimmune) thyroid disease not associated with either malignancy or specific genetic syndromes such as Multiple Endocrine Neoplasia II. Patient age at onset, disease type and gender were recorded.
Comparison of prevalences between the MS and General populations were performed both before and after direct age and sex standardization of population samples. This minimized bias due to different population distributions.
Two-tailed t tests were used for comparison of average age of onset of disease. Chi squared tests were used to compare proportions unless samples were less than 5, and Fisher's exact test was used. An α of 0.05 was used for significance.
Results
A previous MS prevalence study found 493 living patients with MS in NL [11]. In October 2003, 438 addresses were confirmed for a list of living MS patients [11] and these patients were mailed surveys. After one survey mailing and a followup mailing for non-responders, 328 surveys were returned (a 75% rate of return). The results of this survey are shown in Table 2. There were no significant differences in demographics between the patients that returned the surveys and the MS patients in the database.
Table 2 Patient characteristics of survey results.
All MS Patients Survey Returned Significance
Average Age 45.7 45.2 p = 0.53
Female to Male Ratio 2.72:1 2.86:1 p = 0.76
Age of First Symptoms 32.2 32.4 p = 0.81
Proportion RRMS 0.83 0.83 p = 0.97
RRMS – Relapsing Remitting Multiple Sclerosis
532 patients with headache and various injuries were found by chart review. The average age was 39.6 years and the female to male ratio is 2.57:1. There were 274 patients with migraines, 25 people with MVAs, 27 with injuries and the rest were of various other headache types. The prevalences of thyroid disease in both the MS population and in our representative sample of the general population are shown in Table 3. There was an increased prevalence of both Graves' disease and Hashimoto's disease in the MS population compared to the general population, although the co-occurrence of Hashimoto's was not significant after age and sex standardization (Table 3).
Table 3 Prevalence of thyroid disease in MS patients and in a representative sample of the general population
MS Population
(N = 491) General Population
(N = 532) Significance after age-sex standardization
Graves' 15 (3.1%) 2 (0.4%) p = 0.002*
Hashimoto's 26 (5.5%) 12 (2.2%) p = 0.097
* by Fisher exact test
Access to all provincial records made laboratory confirmation of TSH results through electronic and hospital records thorough. Of the 532 patients used for the representative general population, 472 patients had recently TSH results drawn within the past 9 years, most (433 – 81.4%) within 1 year of their neurology consultation or since. In fact, many have had almost yearly serial TSH tests performed for the past several years. We are unable to speculate why this test appears to be so well-used in this general patient population. Electronic and hospital records were also used to confirm the use of thyroid replacement medications as well as the results of ultrasound and nuclear medicine tests for diagnostic workup of thyroid disease. Records of patient visits to endocrinologists were also available for this confirmation.
From the survey results, the chart reviews and the electronic record searches, 15 patients with Graves' disease and 26 patients with Hashimoto's disease were found in the MS population (Table 4). The age of onset of thyroid disease was confirmed for all MS patients with concomitant dysthyroidism. Neither the age of onset of thyroid disease in the MS and general population nor the male to female ratio were significantly different.
Table 4 Demographics of MS patients with thyroid disease and demographics of patients with thyroid disease. Confirmed age means the number of MS patients whose age at onset of thyroid disease could be confirmed.
MS with Graves' Graves' Disease Significance MS with Hashimoto's Hashimoto's Significance
Total Number 15 80 26 57
Confirmed Age 10 13
Average Age of Onset of Thyroid Disease 35.2 37.5 p = 0.61 36.2 37.9 p = 0.70
Female:Male 13:2 72:8 p = 0.66* 21:5 48:9 p = 0.63
* by Fisher exact test
The age of onset (age at appearance of first symptoms) of MS was confirmed for all MS patients with concomitant thyroid disease and for all MS patients. The differences between the age of onset of all MS patients and those MS patients with both Graves' (p = 0.81) and Hashimoto's (p = 0.71) was not significant.
Only 1 out of the 10 MS patients with Graves' disease that responded completely on the survey had undertaken treatment with an interferon-β prior to developing dysthyroidism which she had developed at 56 years of age, 1 year after commencing therapy.
Discussion
Multiple sclerosis (MS) is a chronic inflammatory disease of the central nervous system. It damages both oligodendroglia and axons and may cause paralysis, sensory disturbances, incoordination, visual impairment, and alterations in autonomic and sexual function [15]. Its precise etiology has not been elucidated but many observations suggest that both genetic susceptibility and environmental factors contribute [16]. There continues to be a debate regarding the eligibility of MS as an autoimmune disease [17,18], although it is generally accepted as such [19].
Hashimoto's thyroiditis is an organ-specific autoimmune disease that is the most common cause of goitrous hypothyroidism[20]. Identification as an autoimmune disease follows observations that lymphocytes infiltrate the thyroid [21] and autoantibodies against thyroglobulin, thyroid peroxidase, and (rarely) thyroid hormone stimulating receptor are found in patients with HT[20]. The initiation of the autoimmune events of HT are not precisely known, but could be caused by a molecular mimicry mechanism, abnormal antigen-specific induction of T-cells due to abnormal HLA-related SPC genes, mutation of T cells to form abnormal clones, or an immune defect causing reduced induction of T suppressor cells by specific antigens [22].
Graves' disease is an autoimmune disorder characterized by the production of thyroid-stimulating IgG immunoglobulins directed towards the thyroid simulating hormone (TSH) receptor of the thyroid gland [23]. The exact mechanism for the stimulation of these autoantibodies is unknown, but several have been proposed [24].
There is an increased risk of autoimmune thyroid disease in patients with MS in Newfoundland and Labrador. The study was comparably large (493 MS patients, 532 controls) and there was an excellent survey response rate (75%). There were no gender or age discrepancies in those that responded to the survey (Table 2). Diagnostic confirmation was extensive and case ascertainment was comprehensive. Attempts to reduce reporting and ascertainment bias were made.
Community surveys of hyperthyroidism due to Graves' disease have reported prevalence rates of less than 1.9%, but the prevalence can be as high as 2.7% if subclinical cases are included [13]. Our study showed a rate of 0.38% for Graves' disease (95% CI: 0.07–2.1%) which is compatible with previously reported values. The prevalence of hypothyroidism in other populations may be between 0.2% and 3% [13,25]. Our study showed a rate of 2.3% (95% CI: 1.1–4.9%) in patients with MS, which is again similar to those values reported elsewhere.
Other studies have demonstrated a high prevalence of Graves' disease in first degree relatives of patients with MS [2,7] but not with MS probands (in contrast to our study). Since some MS patients treated with interferon-β may develop laboratory signs of autoimmune thyroid disease [8], it is important to determine whether disease modifying drugs have caused this increased prevalence of Graves' disease in our MS population.
Our survey asked questions relating to previous treatment with disease modifying drugs and their initiation. Only 1 out of the 10 MS patients with Graves' disease that responded completely on the survey had undertaken treatment with an interferon-β prior to developing dysthyroidism which she had developed at 56 years of age, 1 year after commencing therapy. It is unknown whether she would have developed it without the interferon-β, but we can test for significance while excluding her since it is appropriate to look for co-occurrence once other avenues of co-occurrence are factored out[5]. If one only includes the 9 complete respondents via survey who either had not even undertaken interferon-β therapy (N = 5) or had begun therapy after the development of Graves' (N = 4), there is still a significant proportion of MS patients with Graves' disease (p = 0.024).
Although not as significant as the Graves' and MS co-occurrence, the co-occurrence of MS with HT showed a trend towards significance (p = 0.097). This confirms the findings in two previous studies, both of them using laboratory results to show an increase in serum TSH and decrease in T3 and one also using serum autoantibodies [3,6] although other large studies have failed to show an increased risk of other autoimmune diseases (in particular, HT) in multiple sclerosis patients [2,26,27]. The retrospective case ascertainment of HT is most likely subject to greater error than in the case of Graves' disease. Most patients that returned surveys stated that they had primary hypothyroidism, but did not state they had Hashimoto's disease (Table 1). Most charts stated the same. The surveys and charts of subjects with Graves' disease more specifically stated Graves' disease.
Our results suggest a link between autoimmune thyroid disease and MS. There does not seem to be evidence for a temporal link between the two conditions as the age of onset of thyroid disease is the same in the general population as in the MS population. As well, the age of onset of MS is the same in subjects with and without thyroid disease. Therefore, this study does not provide evidence that the onset of one disease is the harbinger of another.
The biological plausibility of an autoimmune pathophysiological link between MS and HT or Graves' disease is unproven. Speculatively, MS is a disease characterized by activated T cells. Activated T cells produce a milieu of cytokines, notably IFN-γ. IFN-γ has been hypothesized to induce the autoimmune process observed in Hashimoto's disease. Therefore, the increased availability of activated T cells in MS may cause an increase of Hashimoto's disease in MS patients. The link between Graves' disease and MS is less clear, however experimental observations using Campath-1H may provide clues. Campath-1H is a molecule that targets the CD52 antigens found on lymphocytes and monocytes. It causes the immune response to change from the Th1 phenotype, thus suppressing MS disease activity [28] but permitting the generation of antibody-mediated thyroid autoimmunity. A significant increase in the rate of development of Graves' disease in MS patients was noted after treatment with Campath-1H [28], but Graves' disease has not been reported in over 600 patients treated with Campath-1H for various other autoimmune disorders including rheumatoid arthritis and psoriasis [28]. Further study along this direction may reveal the reason for the increased co-occurrence between Graves' disease and Multiple Sclerosis.
Conclusion
Hashimoto's disease, Graves' disease and multiple sclerosis are autoimmune disorders that may share similar mechanisms for the pathogenesis of the dysregulation of their self immunity. This study has shown that there is a significant co-occurrence in patients with MS and Graves' disease, and a marginal co-occurrence in patients with MS and HT disease. There appears to be no predilection for earlier age of onset for either MS or the thyroid disorders and there appears to be no gender preference in the co-occurrence of the diseases. Reasons for their increased co-occurrence are discussed.
Acknowledgements
The authors would like to thank Dr Simon Broadley, Gwen Alcock RN of the NL MS Clinic, Don McDonald from the Newfoundland and Labrador Center for Health Information, Dr Blair Fleming and Coleen Owens from the NL Medical Care Plan, Dr Sharon Buehler and Dr V Gadag of the Division of Community Health, MUN, Ken Mullaly of the Government of Newfoundland and Labrador, Vital Statistics Division, and the NL neurologists, both current and previously-practicing.
==== Refs
Davidson A Diamond B Autoimmune diseases N Engl J Med 2001 345 340 350 11484692 10.1056/NEJM200108023450506
Broadley SA Deans J Sawcer SJ Clayton D Compston DA Autoimmune disease in first-degree relatives of patients with multiple sclerosis. A UK survey Brain 2000 123 ( Pt 6) 1102 1111 10825350 10.1093/brain/123.6.1102
Karni A Abramsky O Association of MS with thyroid disorders Neurology 1999 53 883 885 10489063
Roquer J Escudero D Herraiz J Maso E Cano F Multiple sclerosis and Hashimoto's thyroiditis J Neurol 1987 234 23 24 3819783 10.1007/BF00314004
Sloka S Observations on Recent Studies Showing Increased Co-occurrence of Autoimmune Diseases J Autoimmun 2002 18 251 257 12126638 10.1006/jaut.2002.0588
Niederwieser G Buchinger W Bonelli RM Berghold A Reisecker F Koltringer P Archelos JJ Prevalence of autoimmune thyroiditis and non-immune thyroid disease in multiple sclerosis J Neurol 2003 250 672 675 12796827 10.1007/s00415-003-1053-9
Heinzlef O Alamowitch S Sazdovitch V Chillet P Joutel A Tournier-Lasserve E Roullet E Autoimmune diseases in families of French patients with multiple sclerosis Acta Neurol Scand 2000 101 36 40 10660150 10.1034/j.1600-0404.2000.101001036.x
Rotondi M Oliviero A Profice P Mone CM Biondi B Del Buono A Mazziotti G Sinisi AM Bellastella A Carella C Occurrence of thyroid autoimmunity and dysfunction throughout a nine-month follow-up in patients undergoing interferon-beta therapy for multiple sclerosis J Endocrinol Invest 1998 21 748 752 9972674
Sloka JS Pryse-Phillips WE Stefanelli M Multiple sclerosis in Newfoundland and Labrador--a model for disease prevalence Can J Neurol Sci 2005 32 43 49 15825545
Rahman P Jones A Curtis J Bartlett S Peddle L Fernandez BA Freimer NB The Newfoundland population: a unique resource for genetic investigation of complex diseases Hum Mol Genet 2004 13 1287 10.1093/hmg/ddh143
Sloka JS Pryse-Phillips WE Stefanelli M Incidence and prevalence of multiple sclerosis in Newfoundland and Labrador Can J Neurol Sci 2005 32 37 42 15825544
Escobar-Jimenez F Fernandez-Soto ML Luna-Lopez V Quesada-Charneco M Glinoer D Trends in diagnostic and therapeutic criteria in Graves' disease in the last 10 years Postgrad Med J 2000 76 340 344 10824047 10.1136/pmj.76.896.340
Tunbridge WM Evered DC Hall R Appleton D Brewis M Clark F Evans JG Young E Bird T Smith PA The spectrum of thyroid disease in a community: the Whickham survey Clin Endocrinol (Oxf) 1977 7 481 493 598014
Shank JC A study of thyroid disease in family practice J Fam Pract 1976 3 247 253 993751
Paty DW Ebers GC Multiple Sclerosis Contemporary Neurology Series 1998 50 Chapter 4 - Diagnosis
Noseworthy JH Lucchinetti C Rodriguez M Weinshenker BG Multiple sclerosis N Engl J Med 2000 343 938 952 11006371 10.1056/NEJM200009283431307
Chaudhuri A Behan PO Multiple sclerosis is not an autoimmune disease Arch Neurol 2004 61 1610 1612 15477520 10.1001/archneur.61.10.1610
Weiner HL Multiple sclerosis is an inflammatory T-cell-mediated autoimmune disease Arch Neurol 2004 61 1613 1615 15477521 10.1001/archneur.61.10.1613
Roach ES Is multiple sclerosis an autoimmune disorder? Arch Neurol 2004 61 1615 1616 15477522 10.1001/archneur.61.10.1615
Huang W Kukes GD Hashimoto's thyroiditis: an organ-specific autoimmune disease-- pathogenesis and recent developments Lab Invest 1999 79 1175 1180 10532582
Marino M Latrofa F Barbesino G Chiovato L Pathogenetic and clinical aspects of autoimmune thyroiditis Exp Clin Endocrinol Diabetes 1999 107 Suppl 3 S79 S83 10522811
Volpe R LE B Autoimmune thyroid diseases Diseases of the thyroid 1997 Totowa, Humana Press 125 154
McKenna TJ Graves' disease Lancet 2001 357 1793 1796 11403836 10.1016/S0140-6736(00)04906-0
Kohn LD Napolitano G Singer DS Molteni M Scorza R al. Graves' Disease: A Host Defense Mechanism Gone Awry Intern Rev Immunol 2000 19 633 664
Falkenberg M Kagedal B Norr A Screening of an elderly female population for hypo- and hyperthyroidism by use of a thyroid hormone panel Acta Med Scand 1983 214 361 365 6660045
Wynn DR Rodriguez M O'Fallon WM Kurland LT A reappraisal of the epidemiology of multiple sclerosis in Olmsted County, Minnesota Neurology 1990 40 780 786 2082944
De Keyser J Autoimmunity in multiple sclerosis Neurology 1988 38 371 374 3347339
Coles AJ Wing M Smith S Coraddu F Greer S Taylor C Weetman A Hale G Chatterjee VK Waldmann H Compston A Pulsed monoclonal antibody treatment and autoimmune thyroid disease in multiple sclerosis Lancet 1999 354 1691 1695 10568572 10.1016/S0140-6736(99)02429-0
|
16280086
|
PMC1308850
|
CC BY
|
2021-01-04 16:38:03
|
no
|
J Autoimmune Dis. 2005 Nov 9; 2:9
|
utf-8
|
J Autoimmune Dis
| 2,005 |
10.1186/1740-2557-2-9
|
oa_comm
|
==== Front
J Inflamm (Lond)Journal of Inflammation (London, England)1476-9255BioMed Central London 1476-9255-2-141625963310.1186/1476-9255-2-14ResearchDifferential signaling mechanisms regulate expression of CC chemokine receptor-2 during monocyte maturation Phillips Roderick J [email protected] Marin [email protected] Brett [email protected] Department of Physiology David Geffen School of Medicine at University of California, Los Angeles, Los Angeles, California, 90095 USA2 Jonsson Comprehensive Cancer Center, David Geffen School of Medicine at University of California, Los Angeles, Los Angeles, California, 90095 USA3 Department of Discovery Research, Intermune, 3280 Bayshore Blvd, Brisbane, California, 94005 USA4 Department of Technology Development, ChemoCentryx Inc., 1539 Industrial Road, San Carlos, California USA2005 31 10 2005 2 14 14 15 12 2004 31 10 2005 Copyright © 2005 Phillips et al; licensee BioMed Central Ltd.2005Phillips et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Peripheral blood monocytes and monocyte-derived macrophages are key regulatory components in many chronic inflammatory pathologies of the vasculature including the formation of atherosclerotic lesions. However, the molecular and biochemical events underlying monocyte maturation are not fully understood.
Methods
We have used freshly isolated human monocytes and the model human monocyte cell line, THP-1, to investigate changes in the expression of a panel of monocyte and macrophage markers during monocyte differentiation. We have examined these changes by RT-PCR and FACS analysis. Furthermore, we cloned the CCR2 promoter and analyzed specific changes in transcriptional activation of CCR2 during monocyte maturation.
Results
The CC chemokine receptor 2 (CCR2) is rapidly downregulated as monocytes move down the macrophage differentiation pathway while other related chemokine receptors are not. Using a variety of biochemical and transcriptional analyses in the human THP-1 monocyte model system, we show that both monocytes and THP-1 cells express high levels of CCR2, whereas THP-1 derived macrophages fail to express detectable CCR2 mRNA or protein. We further demonstrate that multiple signaling pathways activated by IFN-γ and M-CSF, or by protein kinase C and cytoplasmic calcium can mediate the downregulation of CCR2 but not CCR1.
Conclusion
During monocyte-to-macrophage differentiation CCR2, but not CCR1, is downregulated and this regulation occurs at the level of transcription through upstream 5' regulatory elements.
HumanCellular DifferentiationCell Surface MoleculesGene Regulation
==== Body
Background
Chemokines are a superfamily of small (8–10 kDa) proteins, which coordinate cellular responses to inflammation, insult or injury [1-4]. They also play a pivotal role in the regulation of leukocyte trafficking and extravasation through the luminal surface of endothelial cells into sites of tissue inflammation. The chemokine superfamily includes at least 20 receptors and more than 50 ligands [1-5]. The chemokine ligands can be separated into two major categories depending on whether they express a CC or CXC amino acid motif in their N-termini. This dichotomy appears to be functionally important since many CC chemokines preferentially target monocytes and T cells, while CXC chemokines such as IL-8 (CXCL8) tend to attract neutrophils. The CC chemokines bind to a family of G-protein coupled serpentine (seven transmembrane spanning) receptors, which are termed CC chemokine receptors (CCRs; [1,3,6]). Currently ten of the CC receptors have been identified and monocytes predominantly express three of them: CCR1, CCR2 and CCR5 [2,7,8]. These receptors can bind and signal to different CC chemokines including MCP-1 (CCL2), MIP-1α (CCL3) and RANTES (CCL3) [3,4,9] and these same chemokines are secreted by endothelial cells when activated by LDL or inflammatory cytokines [10-13] or when the endothelium is damaged [14,15].
Indeed, the recruitment of peripheral blood monocytes to the site of injured endothelium by pro-inflammatory chemokines is a key regulatory component in the formation of an atherosclerotic lesion [16,17]. The monocytes subsequently adhere to the endothelium and eventually migrate into the sub-intima [18,19]. Here, they receive a series of differentiation signals including macrophage-colony stimulating factor (M-CSF) and minimally oxidized LDL that enables them to mature into macrophages. These macrophages then engulf large quantities of cholesterol to become lipid-laden foam cells. And it is the accumulation of these foam cells that eventually leads to the formation of characteristic fatty streaks, intermediate lesions and fibrous plaques [20,21].
To date, though, the actual role of chemokines and their receptors in atherosclerosis has not been clearly established. However, recent studies using transgenic mouse models of atherosclerosis have provided convincing evidence that CCR2 is required for disease progression in apolipoprotein E-null mice [22,23]. In these animals, disruption of the CCR2 gene greatly decreases lesion formation without affecting plasma lipid or lipoprotein concentrations. Using a slightly different approach Rollins and colleagues have demonstrated that CCL2, the ligand for CCR2, plays an equally important role in the development of atherosclerosis in low-density lipoprotein receptor deficient mice [24,25]. Here, deletion of CCL2 leads to a significant reduction in lipid deposition within the aorta.
Despite the promising experimental results from these systems, relatively little is known about how the expression of chemokine receptor genes is regulated in normal or diseased human tissues. A recent paper by Yamamoto and colleagues [26] examined the basal regulatory mechanisms underlying expression of the CCR2 gene in the human monocyte cell line, THP-1. Indeed, this group characterized two key elements that seemed to be necessary and sufficient for the basal regulation of CCR2 expression: an Oct-1 binding sequence located 36 bp upstream of the TATA box and a tandem CAAT/enhancer-binding protein (C/EBP) binding sequence located, unusually, in the 5' UTR (at +50 to +77 bp). However, studies have not directly examined the molecular mechanisms by which basal expression of CCR2 is rapidly downregulated during the differentiation of monocytes into macrophages.
In an effort to address this issue, we have further developed a model of monocyte differentiation using THP-1 cells, which can be induced to mature into macrophages using either phorbol esters and ionomycin or a physiological combination of interferon-γ (IFN-γ) and M-CSF. In common with other studies, we report here that THP-1 cell maturation mediated by either high concentrations of PMA (50 nM) alone, or very low concentrations of PMA (1 nM) plus ionomycin (1 μM) is characterized by an increase in size, the development of an adherent phenotype and the up-regulation of a panel of differentiation markers [27-30]; in addition, CCR2, but not CCR1, was specifically down-regulated during differentiation. Modulation of CCR2 by PMA (50 nM), but not PMA (1 nM) plus ionomycin (1 μM), was found to be sensitive to inhibition by the broad-spectrum protein kinase inhibitor staurosporine. Furthermore, transient transfection of THP-1 cells with a CCR2-specific reporter construct indicated that PMA (50 nM) and PMA (1 nM) plus ionomycin (1 μM) mediated the downregulation of CCR2 through inhibition of CCR2-specific gene transcription. Moreover, physiological treatment of THP-1 monocytes with two known differentiation factors, IFN-γ and M-CSF, also promoted a differentiation phenotype essentially identical to that observed using pharmacologic stimuli. These data indicate that the activation of several intracellular signaling pathways selectively regulate the expression of CCR2 during monocyte maturation into macrophages.
Materials and methods
Cell lines
The THP-1 human monocytic cell line (ATCC) was grown in RPMI 1640 medium (GibcoBRL) containing 10 % fetal calf serum (FCS; GibcoBRL), 100 U/ml penicillin and 100 μg/ml streptomycin (GibcoBRL). The cells were maintained in culture at 37°C and 5% C02. Typically, cells (7 × 106 per point) were stimulated with 50 nM phorbol myristate acetate (PMA; Sigma) or 1 nM PMA plus 1 μM ionomycin (Calbiochem) in the presence or absence of the PKC inhibitor staurosporine (Calbiochem).
Isolation and culture of human peripheral blood monocytes
Peripheral blood mononuclear cells (PBMCs) were isolated from freshly prepared leukopacks (buffy coats) that were between 2–4 hours old. Briefly, 20 ml of blood from leukopacks were diluted using PBS (1:1) and layered over 15 ml of Ficoll-Paque PLUS (Amersham Pharmacia Biotech). Cells were then centrifuged at 400 × g for 20 minutes at room temperature. After this time, PBMCs were collected from the interphase and washed (× 2) with PBS and centrifuged at 150 × g for 10 minutes. Monocytes were further isolated from PBMCs using Percoll (Amersham Pharmacia Biotech) gradient centrifugation as previously described [31]. Lipid staining of the monocytes revealed that their purity was greater than 90%. Finally, the cells were resuspended and cultured at 106/ml in RPMI 1640 supplemented with 10% autologous serum, penicillin and streptomycin (GibcoBRL).
Cloning the CCR2 promoter
A 1335 bp fragment of the promoter from the hCCR2 gene was cloned into the pGL3 vector (Promega) using sequences determined by Yamamoto and colleagues [26]. This construct, termed pGL3-1335, contained the tandem C/EBP sites plus 1220 bp of the promoter sequence 5' of the transcriptional start site. The 5' primer contained a restriction site for kpnI, while the 3' primer contained a HindIII site. Each primer started with a 2 bp GC-rich clamp. The full primer sequences used are as follows:
pGL3-1335 5' CGGGTACCGCTGCTTTAGGTCCATTTACCCTC
pGL3-1335 3' GCAAGCTTATTGTACATTGGGTTGAGGTCTCC.
The genomic PCR was performed using an annealing temperature of 55°C (30 seconds) and an extension temperature of 72°C (2 minutes); 30 cycles of PCR were performed.
RNA isolation and RT-PCR
Total RNA was isolated using TRIzol (Life Technologies) and by following the manufacturer's instructions. Briefly, cells were lyzed in TRIzol and then mixed with chloroform. The lysate was then centrifuged to separate RNA, DNA and protein. Total RNA, which is contained in the upper aqueous phase was recovered and mixed with isopropanol to precipitate the RNA. The RNA was finally washed in 75% ethanol to remove impurities and dissolved in water.
5 μg of RNA prepared in this way was then taken and DNase treated to remove further enzymatic contamination, before being reverse transcribed to cDNA using a ProSTAR First Strand RT-PCR kit from Stratagene and by following the manufacturer's instructions.
Subsequently, RT-PCR was performed under standard conditions using primers specific for CCR1, CCR2 and GAPDH. The primer sequences used here were:
CCR1 sense 5'GAAACTCCAAACACCACAGAGGAC
CCR1 antisense 5'TTCGTGAGGAAAGTGAAGGCTG
CCR2 sense 5'CCACATCTCGTTCTCGGTTTATCAG
CCR2 antisense 5'CGTGGAAAATAAGGGCCACAG
CCR3 sense 5'CACTAGATACAGTTGAGACCTTTGG
CCR3 antisense 5'GGTAAAGAGCACTGCAAAGAGTC
CCR4 sense 5'ACCCCACGGATATAGCAGATACC
CCR4 antisense 5'CGTCGTGGAGTTGAGAGAGTACTTG
CCR5 sense 5'GGAGCCCTGCCAAAAAATC
CCR5 antisense 5'CTGTATGGAAAATGAGAGCTGC
CCR6 sense 5'TGGCAAGGGGTATAATTTGGG
CCR6 antisense 5'GACAGTCTGGTACTTGGGTTCACAG
CCR7 sense 5'AGACAGGGGTAGTGCGAGGC
CCR7 antisense 5'GGATGGAGAGCACTGTGGCTAG
CCR8 sense 5'ACCTCAGTGTGACAACAGTGACCG
CCR8 antisense 5'ACCATCTTCAGAGGCCACTTGG
CCR9 sense 5'CACTGAGGATGCCGATTCTGAG
CCR9 antisense 5'CGAAATCTGCGTGGCAGTTC
CXCR1 sense 5'CAGATCCACAGATGTGGGA
CXCR1 antisense 5'GTTTGGATGGTAAGCCTGG
CXCR2 sense 5'AACATGGAGAGTGACAGC
CXCR2 antisense 5'GATGAGTAGACGGTCCTTC
CXCR3 sense 5'TCCTTGAGGTGAGTGACCA
CXCR3 antisense 5'GTATTGGCAGTGGGTGGCG
CXCR4 sense 5'AGTATATACACTTCAGATAAC
CXCR4 antisense 5'CCACCTTTTCAGCCAACAG
CXCR5 sense 5'CTGGACAGATTGGACAACTA
CXCR5 antisense 5'CATCACAACAACTCCCTGA
GAPDH sense 5'TCCATGACAACTTTGGTATCG
GAPDH antisense 5'GTCGCTGTTGAAGTCAGAGGA
The annealing temperature used for RT-PCR was 55°C for 30 seconds and the extension temperature was 72°C for 1 minute; typically 30 cycles of PCR were performed. Under these conditions the product sizes for CCR1, CCR2 and GAPDH were 567 bp, 580 bp and 420 bp respectively.
Antibody staining and FACS analysis
THP-1 cells or PBMCs were resuspended in ice-cold staining buffer (PBS + 2% FCS + 0.1% sodium azide) and incubated with Fc block (Miltenyi Biotec) for 5 minutes at 4°C. Subsequently, primary antibodies were added (anti-CCR1, CCR2, CCR5, CCR7, CXCR2 and CXCR4; R&D Systems) at a final concentration of 0.5 μg/μl. The cells were then incubated at 4°C for 25 minutes, after which time they were washed twice in staining buffer. The secondary antibody used for these experiments was Alexa 488 (Molecular Probes) at a final concentration of 1 μg/μl. This time the cells were incubated at 4°C for 25 minutes in the dark. Following incubation with the secondary antibody, the cells were again washed twice, and then resuspended in 500 μl of staining buffer. Samples were finally analyzed on a FACScan flow cytometer (Becton Dickinson) using Cellquest 3.2.1f1 software. Peripheral blood monocytes, monocyte-derived macrophages and THP-1 cells were also stained for CD36, CD11b and CD68 (all purchased from BD Biosciences).
Transient transfection using DEAE/Dextran
THP-1 cells, grown to a density of 5–8 × 105/ml, were resuspended in Tris-buffered saline (TBS; 25 mM Tris.Cl, pH7.4, 137 mM NaCl, 5 mM KCl, 0.6 mM Na2 HPO4, 0.7 mM CaCl2 and 0.5 mM MgCl2). THP-1 cells (7 × 106 per point) were then added to 1 ml of TBS containing 5 μg of the CCR2 promoter-luciferase construct, 2 μg of the renilla control construct (pRL-SV40; Promega) and 500 μg/ml DEAE/Dextran (final concentration). This mixture was then left at room temperature for one hour. Next, DMSO was added to the cells drop-wise to a final concentration of 10% and incubated for 2 minutes at room temperature. Subsequently, the cells were washed twice in TBS, once in RPMI 1640 medium lacking FCS and antibiotics and once in RPMI 1640 complete medium. The cells were then resuspended in RPMI 1640 complete medium, stimulated with PMA and ionomycin (at the concentrations indicated) and finally incubated at 37°C and 5% CO2 for 48 hours.
After the 48-hour incubation period, cell extracts were made using the luciferase reporter lysis buffer (Promega). Each lysate was subsequently assayed in the dual luciferase reporter assay (Promega) following the manufacturer's instructions. Luciferase activity was determined using a Monolight series 2010 luminometer (Analytical Luminescence Laboratory) and then normalized to the renilla control.
Results
Freshly isolated monocytes selectively downregulate CCR2, but not CCR1, in culture
Human monocytes were isolated from blood leukopacks and placed in culture for up to 5 days (Figure 1). During this time these cells underwent changes in both morphology and gene expression. Freshly isolated monocytes initially appeared small and round, but after 5 days in culture they became adherent, and increased in both size and granularity (Figure 1A). Next, we analyzed changes in the expression of the macrophage differentiation markers CD11b, CD36 and CD68 (Figure 1B). We found that monocytes cultured for 5 days upregulated expression of the integrin CD11b and the scavenger receptors CD36 and CD68, consistent with a change in phenotype from monocyte to macrophage (Figure 1B). Next, we wanted to examine changes in the expression of chemokine receptors as monocytes differentiated into macrophages. Using primers specific for CXCR1-5 and CCR1-CCR9, we performed semi-quantitative analysis of receptor mRNA expression (Figure 1C). Initially, however, we determined the efficacy and specificity of the primers by analyzing genomic DNA samples prepared from freshly isolated monocytes (Figure 1C, panel I). In all cases a single band of the anticipated size was observed indicating that the primers were specific for the desired chemokine receptor. This data further suggested that a lack of chemokine receptor expression observed in freshly isolated monocytes and monocytes cultured for up to five days was a true result, rather than as a reflection of inappropriate primer design. Subsequently, we performed semi-quantitative analysis of receptor mRNA expression on freshly isolated monocytes and monocytes cultured for up to five days (Figure 1C, panel II). Under these conditions, freshly isolated monocytes showed strong expression of CCR1, CCR2, CCR5, CXCR2 and CXCR4 mRNAs, and trace levels of CCR4 and CCR7 mRNA. Expression of CCR1, CCR2, CCR5, and CXCR4 mRNAs remained elevated after two days in culture, while that of CXCR2 decreased and that of CCR7 temporarily increased. However, after five days in culture CCR2 mRNA expression but not that of CCR1, CCR5 or CXCR4 was dramatically downregulated (Figure 1C, panel II). Indeed, levels of CCR5 and CCR1 mRNA actually increased over those observed in freshly isolated monocytes. To confirm the specificity of this effect we subsequently compared cell surface expression of these chemokine receptors in cultured monocytes and freshly isolated monocytes by flow cytometry (Figure 1D). In agreement with our mRNA data, expression of CCR2 protein, but not CCR1, CCR5 and CXCR4 was rapidly downregulated during monocyte maturation. Negligible cell surface expression of CCR7 protein was observed at any of the time points examined, while CXCR2 cell surface expression remained curiously elevated despite downregulation of CXCR2 mRNA, suggesting that the half-life of this protein is actually quite long (Figure 1D).
Figure 1 Macrophage-derived monocytes selectively downregulate CCR2, but not CCR1, during differentiation. (a). Changes in morphology between freshly isolated monocytes (left panel) and cells cultured for 5 days (right panel) were determined using a Nikon Diaphot Camera set up and Axon Imaging Workbench software. Magnification is at 60 ×. (b). Freshly isolated monocytes were either cultured for 5 days (broken line) or immediately stained (solid line) for a panel of macrophage markers: CD36 (left panel), CD11b (middle panel) or CD68 (right panel). Dotted histograms represent the isotype controls. (c). Panel I. Genomic DNA was prepared from freshly isolated monocytes and assayed for germ line expression of chemokine receptors CCR1-CCR9 and CXCR1-CXCR5 by PCR using primers designed in-house. Note each primer pair amplified a single product only, thus confirming that the primers are functional and specific. Panel II. Messenger RNA was prepared from freshly isolated monocytes (upper panel) and cells that had been cultured for either 2 days (middle panel) or 5 days (lower panel). Subsequently, RT-PCR was performed using primers for chemokine receptors CCR1-CCR9, CXCR1-CXCR5 and GAPDH. Marker is a 100 bp DNA ladder. Similar results were obtained in three other experiments. (d). Freshly isolated monocytes (upper panel plots 1, 4, 7, 10, 13 and 16) and cells that had been cultured for either 2 days (middle panel plots 2, 5, 8, 11, 14 and 17) or 5 days (lower panel plots 3, 6, 9, 12, 15 and 18) were stained for CCR1, CCR2, CCR5, CCR7, CXCR2 and CXCR4. Cells were then analyzed for changes in chemokine receptor expression by flow cytometry. Similar results were obtained in three other experiments.
These results indicate that one consequence of monocyte maturation is the selective downregulation of CCR2 gene expression followed by a loss of CCR2 protein from the surface of the cell. While the actual physiological role of this process is unknown, it is likely that CCR2 down-regulation may be involved in restricting 'reverse-migration' of differentiated monocytes back into the blood stream, and thus facilitating capture within the tissues.
PMA-treatment of monocytes induces selective downregulation of CCR2
Based on the above results we decided to further examine the regulation of CCR2 expression in monocyte maturation using the human monocyte cell line, THP-1 and CCR1 as a control. Treatment of these cells with the PKC-activating phorbol ester PMA for 48 hours is a widely accepted procedure for maturing monocytes [27,28]. Cells treated in this way undergo phenotypic changes consistent with their maturation into macrophages [27-30] (also compare Figures 1 and 6).
Figure 6 IFN-γ plus M-CSF promotes a similar differentiation phenotype to that observed using pharmacologic stimuli. (a). THP-1 cells were either left untreated (upper panel) or treated with 500 U/ml IFN-γ plus 5 ng/ml M-CSF (middle panel) or 50 nM PMA (lower panel) for 48 hours. Subsequently, the cells were photographed using a Nikon Diaphot Camera set up and Axon Imaging Workbench software. Magnification is at 40 ×. (b). THP-1 cells were either left untreated or treated for 48 hours with either 50 nM PMA (PMA) or 500 U/ml IFN-γ plus 5 ng/ml M-CSF (I+M) as indicated. Subsequently, these cells were stained with antibodies to macrophage markers CD36 (upper panel), CD11b (middle panel) and CD68 (lower panel) and then analyzed by flow cytometry.
Next, we wanted to determine how treatment of the monocyte cell line, THP-1, with PMA affected the expression of CCR2 in these cells. Thus, monocytes were stimulated with PMA (at the concentrations indicated) for 48 hours and RNA prepared as described above. Our results (Figure 2A) show that CCR2 was selectively down-regulated in a dose dependent manner, whereas expression of CCR1 (the other main CC receptor on monocytes) and the house-keeping gene GAPDH remained unaffected. PMA (50 nM) was sufficient to completely abrogate CCR2 expression (Figure 2A, lane 8), whilst 10 nM PMA reduced expression of this chemokine receptor by approximately 75% (Figure 2A, lane 7). Treatment of THP-1 cells with 1 nM PMA did not affect expression of CCR2 (Figure 2A, lane 6).
Figure 2 PMA induces a dose-dependent selective downregulation of CCR2. (a). THP-1 cells were either untreated (lanes 1, 5 and 9) or treated with PMA at 1 nM (lanes 2, 6 and 10), 10 nM (lanes 3, 7 and 11) or 50 nM (lanes 4, 8 and 12) for 48 hours. Messenger RNA was then prepared and RT-PCR performed using primers for CCR1 (lanes 1–4), CCR2 (lanes 5–8) and GAPDH (lanes 9–12). M is a 100 bp DNA ladder. Similar results were obtained in seven other experiments. (b). THP-1 cells were either left untreated or stimulated with PMA (50 nM) for the times indicated. Subsequently the cells were introduced into a FACScan flow cytometer to measure cell surface expression of CCR1 (left panel) or CCR2 (right panel).
Subsequently, we examined whether PMA modulated the cell surface expression of CCR1 and CCR2 by FACS analysis. THP-1 cells were again stimulated with PMA (50 nM) for the times indicated, before being stained with the appropriate antibodies and then analyzed by flow cytometry (Figure 2B). Whereas the levels of CCR1 remained high throughout the duration of the experiment, CCR2 protein expression decreased dramatically. The majority of the expression was lost by 24 hours and by 48 hours virtually no CCR2 was found on the surface of the cultured THP-1 cells (compare Figure 2B, left and right panels). Thus, THP-1 cells treated with PMA (50 nM) mimics the differentiation process observed in cultured monocytes.
Two distinct signal transduction pathways regulate CCR2 expression during monocyte maturation
Our initial observations suggested that while PMA (50 nM) completely abrogated CCR2 expression, sub-optimal concentrations of this phorbol ester (1 nM) had no effect (Figure 2A). We wondered, therefore, whether the addition of a calcium signal (such as ionomycin) together with the sub-optimal concentration of PMA might provide a sufficiently strong stimulus to affect the expression of CCR2. Thus, we incubated monocytes with PMA (1 nM) and ionomycin at the concentrations indicated for 48 hours, and then analyzed CCR2 expression. Our data indicated that ionomycin alone does not affect expression of CCR2 (Figure 3A, middle panel, lanes 4–6). However, in the presence of a sub-optimal PMA signal (1 nM), there was a selective dose-dependent reduction in CCR2 expression (Figure 3A, middle panel, lanes 7–9). At the same time, similar concentrations of PMA and ionomycin did not affect the levels of CCR1 nor GAPDH (Figure 3A upper and lower panels). Monocytes treated with PMA (1 nM) plus ionomycin (1 μM) were also observed to adopt an adherent phenotype and to increase in size similar to the changes in morphology observed in freshly isolated monocytes (data not shown). Furthermore, cell surface expression of CCR2, but not CCR1, was found to be downregulated in the presence of PMA (1 nM) plus ionomycin (1 μM) after 48 hours (Figure 3B). Thus, sub-optimal concentrations of PMA together with a modest calcium signal combine to mediate a maturation phenotype in monocytes that also includes the selective downregulation of CCR2.
Figure 3 Sub-optimal concentrations of PMA, together with a modest calcium signal, also modulate CCR2. (a). THP-1 cells were either unstimulated (lane1) or treated with PMA 1 nM (lane 2) or 50 nM (lane 3) for 48 hours. Alternatively, the cells were treated with increasing concentrations of the calcium ionophore ionomycin alone (lanes 4–6) or in combination with PMA 1 nM (lanes 7–9) also for 48 hours. Messenger RNA was then prepared and RT-PCR performed using primers for CCR1 (upper panel), CCR2 (middle panel) and GAPDH (lower panel). M represents markers, which are a 100 bp ladder. Similar results were obtained in four other experiments. (b). THP-1 cells were either left untreated or stimulated with PMA (1 nM) and ionomycin (1 μM) for the times indicated. Subsequently the cells were stained for expression of CCR1 (left panel) or CCR2 (right panel) and analyzed by flow cytometry.
To determine whether the selective downregulation of CCR2 observed in PMA versus PMA plus ionomycin treated cells represented the same or two different signaling pathways, we performed an experiment using the broad-spectrum kinase inhibitor, staurosporine (Figure 4). We preincubated THP-1 cells with staurosporine at the concentrations indicated for two hours, and then stimulated with either PMA (50 nM; Figure 4A) or PMA (1 nM) plus ionomycin (1 μM; Figure 4B) for 48 hours. Staurosporine alone (at concentrations up to 200 nM) did not significantly inhibit expression of CCR2 (Figure 4A, lanes 9 and 10 and Figure 4B, lane 6) nor CCR1 (Figure 4A, lanes 3 and 4 and Figure 4B, lane 2). Furthermore, the inhibitor did not abrogate the downregulation of CCR2 mediated by PMA plus ionomycin (Figure 4B, compare lanes 7 and 8). In contrast, staurosporine at 50 nM, but not at 10 nM, blocked the loss of CCR2 in PMA (50 nM) treated cells (Figure 4A, compare lanes 7, 8, 11 and 12).
Figure 4 The PKC-inhibitor staurosporine blocks PMA, but not PMA plus ionomycin, induced downregulation of CCR2. (a). THP-1 cells were either untreated (lanes 1, 2, 7, 8, 13 and 14) or preincubated with 50 nM staurosporine (lanes 3, 5, 9, 11, 15 and 17) or 10 nM staurosporine (lanes 4, 6, 10, 12, 16 and 18) for 2 hours. Subsequently, the cells were stimulated with 50 nM PMA (lanes 2, 5, 6, 8, 11, 12, 14, 17 and 18) for a further 46 hours. Messenger RNA was then prepared and RT-PCR performed using primers for CCR1 (lanes 1–6), CCR2 (lanes 7–12) and GAPDH (lanes 13–18). M is a 100 bp DNA ladder. Similar results were obtained in three other experiments. (b). THP-1 cells were either untreated (lanes 1, 3, 5, 7, 9 and 11) or preincubated with 200 nM staurosporine (lanes 2 and 4, 6 and 8 and 10 and 12) for 2 hours. Subsequently the cells were stimulated with a combination of 1 nM PMA and 1 μM ionomycin (lanes 3 and 4, 7 and 8 and 11 and 12) for a further 46 hours. Messenger RNA was then prepared and RT-PCR performed using primers for CCR1 (lanes 1–4), CCR2 (lanes 5–8) and GAPDH (lanes 9–12). M is a 100 bp DNA ladder. Similar results were obtained in three other experiments.
Thus, these results identify at least two possible signal transduction pathways present in monocytes that could regulate the expression of CCR2 during monocyte differentiation.
CCR2 expression is regulated at the level of transcription
Having established that CCR2 is down-regulated during monocyte differentiation, we next wanted to determine whether the regulation occurs at the level of RNA stability or at the level of transcription. We, therefore, cloned a 1335 bp fragment of the CCR2 promoter using the sequence described by Yamamoto and colleagues [26]. This fragment was then subcloned into the mammalian expression vector pGL3 upstream of the luciferase gene, generating the pGL3-1335 construct. In addition to the sequences upstream of the TATA box, pGL3-1335 included 115 bp of the 5'UTR, which contains the two tandem C/EBP repeats that are thought to be necessary for the basal expression of the CCR2 gene [26].
Subsequently, we transfected this construct into the THP-1 cells using DEAE/dextran and either left the cells untreated, or treated them with PMA (50 nM), or PMA (1 nM) plus ionomycin (1 μM) for 48 hours in the presence or absence of staurosporine (100 nM). Cells were then lyzed and assayed for transcriptional activity. Our results showed that the pGL3-1335 construct, itself, gave a 13-fold induction over the background construct lacking the CCR2 promoter (Figure 5, compare lanes 1 and 2). Furthermore, both PMA and PMA plus ionomycin strongly abrogated this transcriptional activity (Figure 5 lanes 3 and 4) suggesting that the dual signal transduction pathways activated by PMA and PMA plus ionomycin mediated regulation of CCR2 expression at the level of transcription. In the presence of staurosporine, inhibition of CCR2 promoter activity mediated by PMA, but not PMA (1 nM) plus ionomycin (1 μM), was abrogated (Figure 5, compare lanes 6 and 7). Thus, these data indicate that the PMA (but not the PMA plus ionomycin) mediated inhibition of CCR2 promoter activity is ultimately regulated by one or more staurosporine-sensitive transcription factors.
Figure 5 Staurosporine blocks PMA, but not PMA plus ionomycin, induced downregulation of CCR2 promoter activity. THP-1 cells were transfected with either 5 μg of vector alone (pGL3-basic; lane 1) or with 5 μg of the pGL3-1335 construct (lanes 2–7). In addition, each sample was also co-transfected with 2 μg of pRL-SV40 (renilla) to act as an internal control. Cells were then either left untreated (lanes 1–4) or pretreated with staurosporine (100 nM) for 2 hours (lanes 5–7). Next, the THP-1 cells were stimulated with a combination of PMA alone (lanes 3 and 6) or PMA plus ionomycin (lanes 4 and 7) for a further 46 hours. Subsequently, cell extracts were prepared and assayed for both luciferase and renilla activity. After normalization to the renilla control, the CCR2 transcriptional activity was determined relative to the pGL3-basic vector, which was arbitrarily assigned a value of 1. Similar results were obtained in two other experiments
Treatment with IFN-γ and M-CSF produces a similar differentiation phenotype to that seen with PMA and ionomycin
The above results reflect a phenotype induced by pharmacologic agents and we next wanted to ensure that this phenotype is applicable to physiologic agents also. To that end, THP-1 cells treated with IFN-γ plus M-CSF have already been shown to promote monocyte maturation, although it has yet to be confirmed that these agents regulate CCR2 expression at the level of transcription [32]. Initially, though, we wanted to demonstrate that monocytes treated with IFN-γ plus M-CSF showed changes in morphology similar to that observed with freshly isolated monocytes (compare Figures 1 and 6). After 48 hours treatment with IFN-γ plus M-CSF, monocytes became adherent and increased in size similar to that observed for freshly isolated monocytes in culture (compare Figure 1A and Figure 6A middle panel). PMA-treated monocytes also underwent similar changes in morphology (Figure 6A, lower panel). Furthermore, flow cytometric studies revealed that monocytes treated with either IFN-γ plus M-CSF or PMA strongly upregulated the macrophage maturation markers CD11b, CD36 and CD68 (Figure 6B). Similar results were observed for cells treated with PMA plus ionomycin (data not shown). Thus, monocytes treated with a panel of physiologic and pharmacologic stimuli promote maturation to the macrophage phenotype as determined by changes in morphology and upregulation of macrophage maturation markers.
Next, we wanted to determine whether IFN-γ plus M-CSF induced the differentiation-associated downregulation of CCR2 (Figure 7). Therefore, monocytes were treated with IFN-γ (500 U/ml) plus M-CSF (5 ng/ml) for 48 hours and CCR2 mRNA was examined (Figure 7A). Our results showed that IFN-γ plus M-CSF did selectively downregulate CCR2, but not CCR1 in a manner analogous to that observed for PMA and PMA plus ionomycin (Figure 7A upper and middle panels). A similar pattern was also observed when transcriptional activity was examined (Figure 7B). Here, PMA completely down-modulated CCR2 transcription, while the combined effects of IFN-γ plus M-CSF reduced this activity by approximately 70%. In the presence of staurosporine, the inhibition of CCR2 promoter activity mediated by IFN-γ (500 U/ml) plus M-CSF (5 ng/ml) was abrogated in a manner analogous to that observed for PMA (Figure 7B lanes 6 and 7).
Figure 7 IFN-γ plus M-CSF promotes specific down-regulation of CCR2. (a). THP-1 cells were either untreated (lane 1, upper, middle and lower panels) or treated with 500 U/ml IFN-γ plus 5 ng/ml M-CSF (lane 2 upper, middle and lower panels) or 50 nM PMA (lane 3 upper, middle and lower panels) for 48 hours. Messenger RNA was then prepared and RT-PCR performed using primers for CCR1 (upper panel), CCR2 (middle panel) and GAPDH (lower panel). M is a 100 bp DNA ladder. Similar results were obtained in three other experiments. (b). THP-1 cells were transfected with either 5 μg of vector alone (pGL3-basic) or with 5 μg of the pGL3-1335 construct. In addition, each sample was also transfected with 2 μg of pRL-SV40 (renilla) to act as an internal control. Cells were then either left untreated or treated with either 500 U/ml IFN-γ plus 5 ng/ml M-CSF or 50 nM PMA. Subsequently, cell extracts were prepared and assayed for both luciferase and renilla activity. After normalization to the renilla control, CCR2 transcriptional activity was calculated relative to the pGL3-basic vector, which was arbitrarily assigned a value of 1. Similar results were obtained in two other experiments.
Taken together, these data suggest that PMA (50 nM), PMA plus ionomycin and IFN-γ plus M-CSF mediate similar changes in the monocyte phenotype during maturation of these cells. Thus, the monocyte cell line, THP-1, is a useful model system with which to investigate the underlying regulatory mechanisms governing chemokine receptor expression during monocyte differentiation.
Discussion
In this paper we demonstrate that a major consequence of monocyte maturation into macrophages is the selective downregulation of the chemokine receptor, CCR2, but not the related CCR1. We have further shown that there are multiple stimuli, which can selectively down-modulate CCR2 expression, including high concentrations of PMA (50 nM), or low PMA (1 nM) plus ionomycin (1 μM), or IFN-γ (500 U/ml) plus M-CSF (5 ng/ml). Each of these stimuli regulate the expression of CCR2 at the level of transcription, although it appears that at least two different signal transduction pathways are involved based on the ability of staurosporine to interfere with these processes. Treatment of THP-1 monocytes with staurosporine abrogated the ability of PMA and IFN-γ plus M-CSF to downregulate CCR2. By contrast, staurosporine was unable to block PMA plus ionomycin mediated downregulation of CCR2 expression. Thus, this study provides evidence that there is dynamic and selective regulation of the CCR2 gene during monocyte differentiation.
Our results indicate that treatment of THP-1 cells with either PMA alone (50 nM) or PMA (1 nM) plus ionomycin (1 μM) promotes a differentiation phenotype that is characterized by morphological changes and altered CCR2 gene expression. Indeed, these observations have already been noted by other researchers studying monocyte differentiation [27,28,32]. In particular, we show that THP-1 cells rapidly become adherent and their morphology changes from the typical round shape of monocytes to spindle-shaped cells with pseudopodia, which are characteristic of macrophages. At the same time there was also an increase in the size and granularity of the cells. In addition, we demonstrated an up-regulation in expression of genes associated with monocyte differentiation, notably CD11b, CD36 and CD68. Concomitantly, the expression of CCR2, but not CCR1, was selectively downregulated, suggesting that the loss of this chemokine receptor is a consequence of monocyte differentiation. This downregulation was observed at the level of cell surface receptor expression, mRNA expression, and transcription. Clearly, these are specific regulatory events since the levels of CCR1 mRNA are not affected by either combination of pharmacologic agents.
However, when THP-1 cells were treated with PMA (50 nM) or PMA plus ionomycin in the presence of staurosporine, differential results were obtained: PMA-mediated modulation of CCR2 was sensitive to the inhibitory effects of staurosporine (50 nM), whereas staurosporine concentrations as high as 200 nM failed to block PMA plus ionomycin-induced downregulation of CCR2. Staurosporine alone did not promote the loss of either CCR2 or CCR1. These results indicate that staurosporine defines a dichotomy in the regulation of CCR2 expression by PMA (50 nM) versus PMA plus ionomycin that had not previously been appreciated.
Staurosporine, itself, is a broad-spectrum inhibitor of protein kinases including PKA, PKC, and PKG. PMA has classically been shown to act almost exclusively through PKC and this would explain why staurosporine was able to block the PMA-induced downregulation of CCR2. By inference, PMA plus ionomycin would appear to act through a signal transduction pathway that is not inhibited by staurosporine and presumably this means that second messengers other than PKA, PKC and PKG are involved. To that end, calcineurin, a calcium-sensitive phosphatase may be a target for PMA plus ionomycin [33]. An increase in the intracellular calcium concentration (such as that afforded by the presence of ionomycin) promotes a conformational change in calcineurin, which then dephosphorylates and activates the transcription factor NFAT facilitating its translocation to the nucleus. In addition, it has been shown that PMA enhances the calcium sensitivity of NFAT, thus creating a synergistic signal [33,34]. This synergy may result from de novo synthesis and post-translational modification of another transcription factor termed activating protein-1, AP-1 [33,34]. Indeed, NFAT proteins show a characteristic ability to co-operate with AP-1 in DNA-binding and transactivation [33,34]. Interestingly, in the region of the CCR2 promoter that we cloned there are two putative binding sites for AP-1 (core binding motif TGA(C/G)TCA) and three putative binding sites for NFAT (core binding motif GGAAA) as determined by the MatInspecter transcription factor binding site analysis program. It has also been suggested that additional transcription factors including OCT1 and C/EBP can act synergistically with NFAT and again there are multiple binding sites for each of these DNA-binding proteins in the CCR2 promoter, although at this stage we have no evidence to suggest that they are involved in the physiological regulation of CCR2 gene expression.
A requirement for co-operation and cross-talk between these two pharmacologic agents is further supported by the fact that ionomycin alone (at concentrations as high as 1 μM) was unable to down-modulate CCR2.
Some reports have suggested that CCL2 could be involved in the early stages of CCR2 protein down-modulation, while other studies indicate that the differentiation process itself, is a major factor in the selective loss of CCR2 gene expression [8,32]. Numerous cytokines are known to be involved in monocyte activation and differentiation, among them M-CSF and IFN-γ [32,35,36]. M-CSF is a lineage-specific hematopoetic growth factor that stimulates monocyte differentiation [35,36]. The c-fms proto-oncogene encodes a high affinity receptor for M-CSF [37] and it has been shown that THP-1 cells express this protein and that it is up-regulated during differentiation. However, cells stimulated with M-CSF alone for 48 hours did not lose expression of CCR2 (data not shown).
Conversely, IFN-γ alone, which is constitutively expressed by monocyte lineage cells and which promotes maturation of monocytes to macrophages [38], did significantly reduce expression of CCR2, although the cells did not become adherent and neither did they change their morphology (data not shown). Interestingly, IFN-γ has been demonstrated to up-regulate levels of M-CSF in monocytes during maturation [38] and when both IFN-γ and M-CSF were added, THP-1 cells did become adherent, changed their morphology and selectively lost CCR2, but not CCR1 – all of which are characteristics of the monocyte differentiation phenotype. These results are in keeping with the studies published by Tangirala and colleagues, who reported similar phenomena in THP-1 cells [32]. In addition, our studies also demonstrated that the regulatory effects mediated by IFN-γ plus M-CSF occurred at the level of transcription, where a significant down-regulation in CCR2 promoter activity was observed. Moreover, in the presence of staurosporine, IFN-γ plus M-CSF was unable to down-regulate levels of CCR2. This result probably reflects the fact that IFN-γ signals extensively through the JAK-STAT pathway, and studies have suggested that staurosporine can block phosphorylation of Janus kinases [39,40]. In addition, we have found two putative binding sites in the CCR2 promoter for STAT transcription factors which would further support the contention that these transcription factors may be important in the regulation of IFN-γ mediated downregulation of CCR2.
Conclusion
This study demonstrates that expression of the chemokine receptor CCR2 is exquisitely correlated with monocyte maturation. Freshly isolated monocytes express high levels of both CCR2 RNA and protein, whereas monocyte-derived macrophages express neither CCR2 RNA nor protein. Conversely, levels of the closely-related chemokine receptor CCR1 remained stable and elevated throughout monocyte maturation. An analysis of the biochemical and molecular mechanisms underlying the regulated expression of CCR2 revealed the existence of several signaling pathways that selectively down-modulate CCR2 gene expression during monocyte differentiation; this expression was largely regulated at the level of transcription. Signaling through PMA and IFN-γ plus M-CSF, but not PMA plus ionomycin was abrogated by prior treatment of the THP-1 cells with staurosporine. Although the physiological role of this process is not well understood, it is likely that CCR2 down-regulation may be involved in restricting the 'reverse-migration' of differentiated monocytes back into the blood stream. This in turn facilitates the retention of differentiated monocytes within inflamed tissues. Thus, by improving our understanding of the regulatory mechanisms that govern CCR2 expression on monocyte lineage cells, we can better appreciate how monocyte recruitment and activation is controlled during chronic inflammatory pathologies such as atherosclerosis.
Competing interests
Brett Premack is currently employed as the Director of Technology at Chemocentryx. Dr. Premack is the holder of stocks within this company. This company investigates the role of chemokines and their receptors as potential therapeutics. One of these projects is to investigate the role of CCR2 antagonists in cardiovascular disease and a phase I clinical trial is ongoing. At the time this study was undertaken Dr. Premack was an unpaid consultant for Chemocentryx. Neither Roderick Phillips nor Marin Lutz has a competing interest in this work.
Authors' contributions
RJP wrote the manuscript and performed all of the experiments except Figure 1A and 1C. ML performed experiments featured in Figure 1A and 1C. BP conceived of the study, and participated in its design and coordination and helped to draft the manuscript. All authors read and approved the final manuscript.
RJP was supported by the American Heart Association grant 9960044Y
==== Refs
Rossi D Zlotnik A The biology of chemokines and their receptors Annu Rev Immunol 2000 18 217 242 10837058 10.1146/annurev.immunol.18.1.217
Murphy PM Baggiolini M Charo IF Hebert CA Horuk R Matsushima K Miller LH Oppenheim JJ Power CA International union of pharmacology. XXII. Nomenclature for chemokine receptors Pharmacol Rev 2000 52 145 176 10699158
Premack BA Schall TJ Chemokine receptors: gateways to inflammation and infection Nat Med 1996 2 1174 1178 8898734 10.1038/nm1196-1174
Baggiolini M Chemokines and leukocyte traffic Nature 1998 392 565 568 9560152 10.1038/33340
Murphy PM International Union of Pharmacology. XXX. Update on chemokine receptor nomenclature Pharmacol Rev 2002 54 227 229 12037138 10.1124/pr.54.2.227
Power CA Wells TN Cloning and characterization of human chemokine receptors Trends Pharmacol Sci 1996 17 209 213 8763196 10.1016/0165-6147(96)10019-5
Charo IF Myers SJ Herman A Franci C Connolly AJ Coughlin SR Molecular cloning and functional expression of two monocyte chemoattractant protein 1 receptors reveals alternative splicing of the carboxyl-terminal tails Proc Natl Acad Sci U S A 1994 91 2752 2756 8146186
Fantuzzi L Borghi P Ciolli V Pavlakis G Belardelli F Gessani S Loss of CCR2 expression and functional response to monocyte chemotactic protein (MCP-1) during the differentiation of human monocytes: role of secreted MCP-1 in the regulation of the chemotactic response Blood 1999 94 875 883 10419877
Neote K DiGregorio D Mak JY Horuk R Schall TJ Molecular cloning, functional expression, and signaling characteristics of a C-C chemokine receptor Cell 1993 72 415 425 7679328 10.1016/0092-8674(93)90118-A
Brown Z Gerritsen ME Carley WW Strieter RM Kunkel SL Westwick J Chemokine gene expression and secretion by cytokine-activated human microvascular endothelial cells. Differential regulation of monocyte chemoattractant protein-1 and interleukin-8 in response to interferon-gamma Am J Pathol 1994 145 913 921 7943180
Goebeler M Yoshimura T Toksoy A Ritter U Brocker EB Gillitzer R The chemokine repertoire of human dermal microvascular endothelial cells and its regulation by inflammatory cytokines J Invest Dermatol 1997 108 445 451 9077472 10.1111/1523-1747.ep12289711
Marfaing-Koka A Devergne O Gorgone G Portier A Schall TJ Galanaud P Emilie D Regulation of the production of the RANTES chemokine by endothelial cells. Synergistic induction by IFN-gamma plus TNF-alpha and inhibition by IL-4 and IL-13 J Immunol 1995 154 1870 1878 7530744
Martin T Cardarelli PM Parry GC Felts KA Cobb RR Cytokine induction of monocyte chemoattractant protein-1 gene expression in human endothelial cells depends on the cooperative action of NF-kappa B and AP-1 Eur J Immunol 1997 27 1091 1097 9174597
Kumar AG Ballantyne CM Michael LH Kukielka GL Youker KA Lindsey ML Hawkins HK Birdsall HH MacKay CR LaRosa GJ Rossen RD Smith CW Entman ML Induction of monocyte chemoattractant protein-1 in the small veins of the ischemic and reperfused canine myocardium Circulation 1997 95 693 700 9024159
Wysocki SJ Zheng MH Smith A Lamawansa MD Iacopetta BJ Robertson TA Papadimitriou JM House AK Norman PE Monocyte chemoattractant protein-1 gene expression in injured pig artery coincides with early appearance of infiltrating monocyte/macrophages J Cell Biochem 1996 62 303 313 8872602 10.1002/(SICI)1097-4644(199609)62:3<303::AID-JCB1>3.0.CO;2-V
Ross R The pathogenesis of atherosclerosis: a perspective for the 1990s Nature 1993 362 801 809 8479518 10.1038/362801a0
Berliner JA Navab M Fogelman AM Frank JS Demer LL Edwards PA Watson AD Lusis AJ Atherosclerosis: basic mechanisms. Oxidation, inflammation, and genetics Circulation 1995 91 2488 2496 7729036
Cai JP Hudson S Ye MW Chin YH The intracellular signaling pathways involved in MCP-1-stimulated T cell migration across microvascular endothelium Cell Immunol 1996 167 269 275 8603436 10.1006/cimm.1996.0035
Randolph GJ Furie MB A soluble gradient of endogenous monocyte chemoattractant protein-1 promotes the transendothelial migration of monocytes in vitro J Immunol 1995 155 3610 3618 7561060
Ross R Cell biology of atherosclerosis Annu Rev Physiol 1995 57 791 804 7778883 10.1146/annurev.ph.57.030195.004043
Lusis AJ Atherosclerosis Nature 2000 407 233 241 11001066 10.1038/35025203
Boring L Gosling J Cleary M Charo IF Decreased lesion formation in CCR2-/- mice reveals a role for chemokines in the initiation of atherosclerosis Nature 1998 394 894 897 9732872 10.1038/29788
Charo IF Peters W Chemokine receptor 2 (CCR2) in atherosclerosis, infectious diseases, and regulation of T-cell polarization Microcirculation 2003 10 259 264 12851643 10.1038/sj.mn.7800191
Gu L Okada Y Clinton SK Gerard C Sukhova GK Libby P Rollins BJ Absence of monocyte chemoattractant protein-1 reduces atherosclerosis in low density lipoprotein receptor-deficient mice Mol Cell 1998 2 275 281 9734366 10.1016/S1097-2765(00)80139-2
Rollins BJ Chemokines and atherosclerosis: what Adam Smith has to say about vascular disease J Clin Invest 2001 108 1269 1271 11696568 10.1172/JCI200114273
Yamamoto K Takeshima H Hamada K Nakao M Kino T Nishi T Kochi M Kuratsu J Yoshimura T Ushio Y Cloning and functional characterization of the 5'-flanking region of the human monocyte chemoattractant protein-1 receptor (CCR2) gene. Essential role of 5'-untranslated region in tissue-specific expression J Biol Chem 1999 274 4646 4654 9988701 10.1074/jbc.274.8.4646
Tontonoz P Nagy L Alvarez JG Thomazy VA Evans RM PPARgamma promotes monocyte/macrophage differentiation and uptake of oxidized LDL Cell 1998 93 241 252 9568716 10.1016/S0092-8674(00)81575-5
Rovera G Santoli D Damsky C Human promyelocytic leukemia cells in culture differentiate into macrophage-like cells when treated with a phorbol diester Proc Natl Acad Sci U S A 1979 76 2779 2783 288066
Naito M Umeda S Yamamoto T Moriyama H Umezu H Hasegawa G Usuda H Shultz LD Takahashi K Development, differentiation, and phenotypic heterogeneity of murine tissue macrophages J Leukoc Biol 1996 59 133 138 8603984
Yesner LM Huh HY Pearce SF Silverstein RL Regulation of monocyte CD36 and thrombospondin-1 expression by soluble mediators Arterioscler Thromb Vasc Biol 1996 16 1019 1025 8696941
Seager Danciger J Lutz M Hama S Cruz D Castrillo A Lazaro J Phillips R Premack B Berliner J Method for large scale isolation, culture and cryopreservation of human monocytes suitable for chemotaxis, cellular adhesion assays, macrophage and dendritic cell differentiation J Immunol Methods 2004 288 123 134 15183091 10.1016/j.jim.2004.03.003
Tangirala RK Murao K Quehenberger O Regulation of expression of the human monocyte chemotactic protein-1 receptor (hCCR2) by cytokines J Biol Chem 1997 272 8050 8056 9065478 10.1074/jbc.272.12.8050
Rao A Luo C Hogan PG Transcription factors of the NFAT family: regulation and function Annu Rev Immunol 1997 15 707 747 9143705 10.1146/annurev.immunol.15.1.707
Macian F Lopez-Rodriguez C Rao A Partners in transcription: NFAT and AP-1 Oncogene 2001 20 2476 2489 11402342 10.1038/sj.onc.1204386
Lenny N Westendorf JJ Hiebert SW Transcriptional regulation during myelopoiesis Mol Biol Rep 1997 24 157 168 9291089 10.1023/A:1006859700409
Clarke S Gordon S Myeloid-specific gene expression J Leukoc Biol 1998 63 153 168 9468274
Nienhuis AW Bunn HF Turner PH Gopal TV Nash WG O'Brien SJ Sherr CJ Expression of the human c-fms proto-oncogene in hematopoietic cells and its deletion in the 5q- syndrome Cell 1985 42 421 428 4028159 10.1016/0092-8674(85)90099-6
Scheibenbogen C Andreesen R Developmental regulation of the cytokine repertoire in human macrophages: IL-1, IL-6, TNF-alpha, and M-CSF J Leukoc Biol 1991 50 35 42 2056245
Fiorucci G Percario ZA Marcolin C Coccia EM Affabris E Romeo G Inhibition of protein phosphorylation modulates expression of the Jak family protein tyrosine kinases J Virol 1995 69 5833 5837 7637029
Callus BA Mathey-Prevot B Interleukin-3-induced activation of the JAK/STAT pathway is prolonged by proteasome inhibitors Blood 1998 91 3182 3192 9558373
|
16259633
|
PMC1308851
|
CC BY
|
2021-01-04 16:36:23
|
no
|
J Inflamm (Lond). 2005 Oct 31; 2:14
|
utf-8
|
J Inflamm (Lond)
| 2,005 |
10.1186/1476-9255-2-14
|
oa_comm
|
==== Front
J Inflamm (Lond)Journal of Inflammation (London, England)1476-9255BioMed Central London 1476-9255-2-151630955210.1186/1476-9255-2-15ResearchNurr1 dependent regulation of pro-inflammatory mediators in immortalised synovial fibroblasts Davies Mark R [email protected] Christine J [email protected] Stephanie [email protected] Kurt [email protected] Andrew E [email protected] Mark [email protected] Maurice RC [email protected] Respiratory and Inflammation Research Department, AstraZeneca, Mereside, Alderley Park, Macclesfield, Cheshire, SK10 4TG, UK2005 25 11 2005 2 15 15 3 8 2005 25 11 2005 Copyright © 2005 Davies et al; licensee BioMed Central Ltd.2005Davies et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Nurr1 is an orphan member of the nuclear receptor superfamily; these orphan receptors are a group for which a ligand has yet to be identified. Nurr1 has been shown to regulate the expression of a small number of genes as a monomeric, constitutively active receptor. These Nurr1 regulated genes are primarily associated with dopamine cell maturation and survival. However, previous reports have shown an increased expression of Nurr1 in the synovium of patients with rheumatoid arthritis (RA) suggesting a pro-inflammatory role for Nurr1 in RA. In this study we investigate the potential pro-inflammatory role of Nurr1 by monitoring Nurr1 dependent gene expression in an immortalised synoviocyte cell line, K4IM.
Methods
We overexpressed the wild type and a dominant negative form of the orphan nuclear receptor Nurr1, in a model synoviocyte cell line. Using the Affymetrix HG-U133 Genechips we demonstrate the effects on the transcriptome by the receptor. Further evidence of gene expression change was demonstrated using quantitative RT-PCR and ELISA analysis.
Results
We show that Nurr1 regulates transcription of a small number of genes for pro-inflammatory modulators of which the most significant is interleukin-8 (IL-8). We also demonstrate increased synthesis and secretion of IL-8 further supporting a role for Nurr1 in inflammatory signalling pathways.
Conclusion
Using microarray analysis we show that elevated levels of Nurr1 leads to increased gene expression of pro-inflammatory genes: IL-8, Amphiregulin and Kit ligand in a model cell line. This data provides further evidence for an additional role for Nurr1 in inflammation and may play a role in the pathogenesis of rheumatoid arthritis.
==== Body
Background
Nuclear receptors can generally be described as ligand activated transcription factors that form a large superfamily of proteins. In humans 48 such receptors have been identified [1] and are involved with an extensive number of cellular processes throughout development and adult physiology [2]. Nuclear receptors are activated through binding by a diverse range of natural and synthetically produced ligand molecules including hormones, fatty acids and antibiotics. In addition to the receptors known to bind ligand, a group of nuclear receptors exist for which a ligand has not been identified; these are termed the orphan nuclear receptors. Among this group of orphans is Nurr1 (NR4A2), a member of the NR4 group of orphan nuclear receptors together with Nur77 (NR4A1) and NOR-1 (NR4A3) [3]. This family can bind as monomers to DNA response elements in the promoters of genes and activate transcription in the absence of ligand [4]. Interestingly, this family of receptors are also capable of binding as a heterodimer with the 9-cis-retinoic acid receptor, RXR [5] or as a heterodimer with other Nur-family members [6]. RXR as a heterodimer with Nurr1 remains active, suggesting that regulation can be modified by the use of specific rexinoids to enhance the response of these receptors to growth factors and therefore this could provide a novel therapeutic avenue for treatment of Nurr1 regulated disease.
The structure of Nurr1 has recently be solved highlighting differences between Nurr1 and the known liganded nuclear receptors [7]. Based upon homology modelling, the region in Nurr1 that would normally contain the ligand-binding pocket has been shown to be substantially different from other nuclear receptors suggesting that there is insufficient space in the putative ligand binding pocket to accommodate a ligand. This may explain the observations that Nurr1 is able to activate transcription in a ligand independent manner and why no ligand has yet been reported for Nurr1.
In contrast to the majority of nuclear receptors, Nurr1 is encoded by an immediate early gene that is rapidly induced in cells in response to external stimuli such as cytokines. Nurr1 has been implicated in a number of human diseases including Parkinson's disease [8,9], schizophrenia [10], alcohol dependence [11] and rheumatoid arthritis [12]. In accordance with the nature of these diseases, the expression of Nurr1 is observed in the developing and adult CNS and within the inflamed synovium of the rheumatic joint [12-14].
A number of genes have been demonstrated to be regulated by Nurr1; many of these are involved with the development and maintenance of midbrain dopaminergic neurons. These include tyrosine hydroxylase [4], Ret tyrosine kinase [15] and the dopamine transporter (SLC6A3) [16]. The Nur-family members have also been shown to play a pivotal role in regulating expression of CRH and pro-opiomelanocortin (POMC) within the hypothalamic-pituitary-adrenal (HPA) axis [6,17-19]. These studies, carried out in mouse pituitary cells, also demonstrate that CRH is capable of causing increased expression of Nurr1, suggesting the presence of a positive feedback loop. More recently, studies in synovial tissue taken from the rheumatoid joint have shown Nurr1 to be highly expressed. In addition it has been demonstrated that inflammatory cytokines are capable of increasing Nurr1 expression through NF-κB and CREB dependent signalling and this elevation in Nurr1 leads to increased transcription of CRH [12,20]. Whilst in the HPA axis, these close Nur-family members are capable of sharing overlapping roles, in primary synoviocytes, treatment with various cytokines shows predominantly Nurr1 upregulation [20]. Therefore in rheumatic synovium it is proposed that Nurr1 is acting to exacerbate the inflammatory response by driving a Nurr1-CRH positive feedback loop, indicating its potential as a target for possible therapeutic intervention.
The purpose of this study was to further investigate the role of Nurr1 in regulating inflammatory processes in synovial cells, and by using transcript profiling to identify Nurr1 regulated genes in synoviocytes, which may be playing a role in the pathology of rheumatoid arthritis.
Methods
Plasmid expression constructs
A reporter gene construct was generated by ligating oligonucleotides representing 3 tandem repeated consensus Nurr1 binding sites (NurRE), into the SpeI/AflII site of the SW-gal construct as previously described [21], to generate the construct: pNurRE3gal. The sense strand oligonucleotides (5'-3') for the insert were as follows, 3XNurRE: ctagtgtgacctttattctcaaaggtcagtgacctttattctcaaaggtcagtgacctttattctcaaaggtcac. Construction of the control plasmid pCMV/hGH has been previously described [22]. Dominant negative constructs were produced by fusing the DNA binding domain of Nurr1 (aa94–365) and the Drosophila engrailed domain (aa2–298) into the pcDNA3.1 expression vector to give pcDNA3.1-Nurr1-DN. The Nurr1 wild type gene was PCR amplified from whole brain RNA and cloned into the pcDNA3.1-V5-HIS vector and the DNA sequence verified.
Cell culture
K4IM cells (a generous gift from E. Murphy) were cultured in RPMI-1640 media supplemented with 10% (v/v) foetal calf serum (FCS), 2 mM glutamine and 50 μg/ml penicillin-streptomycin (Gibco BRL). HeLa cells were cultured in DMEM media supplemented with 10% (v/v) foetal calf serum (FCS), 2 mM glutamine and 50 μg/ml penicillin-streptomycin (Gibco BRL). For reporter gene assays, HeLa cells were cultured in phenol red-free DMEM supplemented with 0.5% (v/v) FCS, 2 mM glutamine and 50 μg/ml penicillin-streptomycin.
Reporter gene assays
24 hrs prior to transfection, 2.2 × 106 HeLa cells were seeded in a 9 cm2 dish, in full growth medium. Cells were transfected by calcium phosphate precipitation with a total of 20–25 μg DNA per 1 ml of precipitate as previously described [23]. Generally, 10–15 μg of reporter construct was co-transfected with 2–5 μg of Nurr1 or empty vector DNA (pcDNA3.1) and including 0.5 μg CMV/hGH plasmid for data normalisation. 6 hrs post-transfection the cells were exposed to osmotic shock (15% glycerol in DMEM for 90 seconds) and the media replaced with phenol-red free DMEM supplemented with 0.5% FCS (assay medium). Cells were harvested 16–20 hrs later by incubating the cells with trypsin-EDTA (phenol red free) for 3 min at 37°C. Cells were counted, seeded into 96 well microtitre plates at 105 cells/well in 100 μl of assay medium and incubated for a further 16–20 hrs. β-galactosidase activity was measured using the spectrophotometric substrate, CPRG (Boehringer) as previously described [21]. Briefly, 100 μl of a cocktail containing 7 μl 50 mM CPRG, 7 μl Z buffer pH7 (600 mM Na2HPO4; 400 mM NaH2PO4; 100 mM KCl; 10 mM MgSO4; 500 mM β-mercaptoethanol), 1 μl 20% SDS and 85 μl dH2O water was added directly to each well and the plates incubated at 37°C before an OD 570 nm measurement was taken. Incubation times varied depending on the individual assays but were typically 30 min–3 hrs. Secreted growth hormone from the cell supernatants was measured using a 2-antibody sandwich ELISA as previously described [22]. β-galactosidase values were normalised using the hGH values.
Determination of IL-8 protein levels
IL-8 was quantified using ELISA (R&D systems – D8050). K4IM cells were resuspended in Amaxa solution "R" to a concentration of 0.4 × 106 cells per 100 μl. A total of 2 μg of DNA was transfected using the Amaxa Nucleofector following the manufacturer's protocol (A-23). Media was collected 48 hr post nucleofection and used undiluted in duplicate according to the manufacturer's instructions.
Quantitative RT-PCR
TaqMan real-time quantitative polymerase chain reaction (PCR) assay was performed using an ABI Prism 7700 Sequence Detection System, according to the manufacturer's protocol (Applied Biosystems), sequences for primers and probes can be found in Table 1. Additionally, primers and probes for IL-8 were obtained as a pre-formulated 20× mix from Applied Biosystems. Amplification of GAPDH (primer/probe mix 4310884E) was performed to standardize the quantification of target RNA, allowing relative quantitation using the ABI Prism 7700 SDS v1.9 software. Briefly, 25 ng of a mixture of brain, placenta and testis total RNA (Ambion 7962, 7950 & 7972) and subsequent 5-fold serial dilutions down to 1/3125 of neat were amplified in triplicate for both GAPDH and each target gene to produce a standard curve. RNA was extracted from cells using TRIzol reagent (Gibco-BRL) according to the manufacturer's guidelines. Total RNA was analysed and quantified using the Agilent Bioanalyser 2100 with the RNA Nano6000 chip. For the purpose of Taqman and RT-PCR analysis, the RNA was DNase treated using the DNase-Away kit (Ambion) according to manufacturer's protocol. 5 μl of RNA at a concentration of 5 ng/μl was dispensed in triplicate into optical 96-well plates for Taqman RT-PCR. Each sample was supplemented with both respective forward, reverse primer and fluorescent labelled probe in a total reaction volume of 25 μl using Taqman Quantitect Probe Master-Mix and RT enzyme mix (Qiagen – 204443). Each target probe was amplified in a separate 96-well plate. All samples were incubated for an initial reverse transcription reaction at 50°C for 30 minutes and then at 95°C for 15 minutes, followed by 40 cycles at 95°C for 15 seconds and 60°C for 1 minute.
Table 1 Sequences for Taqman RT-PCR. Sequences for primers and probes were designed using Primer Express (Applied Biosystems).
Gene Size of product Primer Sequence (5'-3')
Nurr1 77 bp Forward TGTGTTCAGGCGCAGTATGG
Reverse TCCCGAAGAGTGGTAACTGTAGC
Probe CCTCGCCTCAAGGAGCCAGCC
AREG 70 bp Forward ACTCGGCTCAGGCCATTATG
Reverse AAAATGGTTCACGCTTCCCA
Probe TGCTGGATTGGACCTCAATGACACCTACT
KITLG 75 bp Forward TGGTGGCAAATCTTCCAAAAG
Reverse CAATGACTTGGCAAAACATCCA
Probe CATGATAACCCTCAAATATGTCCCCGGG
DNA microarray
Nucleofection was used to transfect the following three expression constructs into the K41M cell line: 1. pcDNA3.1 blank vector (control); 2. pcDNA3.1-Nurr1-WT (Nurr1 WT); 3. pcDNA3.1-Nurr1194–365EnR2–298 dominant negative (DN) co-transfected with pcDNA3.1-Nurr1-WT (Nurr1 DN/Nurr1 WT). Each transfection was carried out in triplicate, thus generating 9 samples from which RNA was extracted. Total RNA was extracted from each sample using RNeasy kit (Qiagen – 74104). RNA integrity and yield were analysed and quantified using the Agilent Bioanalyser 2100 with the RNA Nano6000 chip prior to synthesis of cRNA probes. Preparation of cRNA, hybridization and scanning of the HG-U133 GeneChip oligonucleotide arrays were performed according to the manufacturer's protocol (Affymetrix, Santa Clara, CA). GeneChip images were quantified and gene expression values were calculated by Affymetrix Microarray suite version 5.0 (Mas 5.0, Affymetrix). For statistical analysis a one-way analysis of variance (ANOVA) was used to compare Nurr1 WT to control samples. Prior to the test, probesets that were absent and/or had an expression signal less than 200 in all samples were removed. Probesets that had P-values less than 0.01 were considered statistically significant. A similar one-way ANOVA analysis across each probeset was used to compare Nurr1 DN to control samples. Candidate genes were retained if they showed greater than 1.5-fold upregulation, (p < 0.01) between the control and Nurr1 WT samples and yet no significant regulation (p < 0.01) between the control and Nurr1 DN samples. These remaining genes were then manually chosen for genes exhibiting an ideal profile (being changed with Nurr1 and returned towards basal activity with Nurr D/N).
Results
Dominant negative Nurr1 blocks Nurr1 induced reporter gene expression
To demonstrate the transcriptional activity of Nurr1 and its subsequent attenuation with dominant negative constructs, we used a reporter gene assay to show induction of the LacZ gene under the control of NurRE's. In HeLa cells transfected with the wildtype Nurr1 expression vector, activation of NurRE driven reporter gene constructs was observed. This is consistent with previous reports showing that in the absence of a ligand this family of receptors are constitutively active when assayed using transfected cell-lines [24,25]. In order to assess the specific role of Nurr1 on gene expression in K4IM cells we used a dominant negative Nurr1, which contained the DNA binding domain of the Nurr1 (2–298) fused to the Drosophila engrailed repressor domain. This type of dominant negative has been used extensively to study the role of transcription factor knockout phenotypes [26,27]. Transfection of K4IM cells with wildtype Nurr1 expression constructs causes induction of expression of the reporter gene whilst co-transfection of the dominant negative with the wild-type Nurr1 attenuates this effect at a range of concentrations (Figure 1). This dominant negative Nurr1 activity was therefore used to assess the specific role of Nurr1 activity in K4IM cells.
Figure 1 Coexpression of a dominant negative-Nurr1 receptor attenuates the transactivation activity of Nurr1-wild type in a β-Gal reporter assay. HeLa cells coexpressing pCDNA3.1-Nurr1-WT and pcDNA3.1-Nurr1-DN in varying ratios demonstrates strong antagonist effects of dominant negative Nurr1 construct on Nurr1 transcriptional activity (pNurRE3gal), values normalised to cotransfected hGH, using an hGH sandwich ELISA assay. Values expressed represent the mean fold change ± SEM, compared to the reporter alone.
Analysis of Nurr1 mediated gene expression in K4IM cells
The synovial fibroblast cell line K4IM was chosen as a model cell line to explore the role of Nurr1 in pro-inflammatory signalling pathways relevant to the arthritic joint [12,28]. Cells were transfected with Nurr1 constructs and RNA extracted from cells 16 hours post nucleofection. This was used to prepare cRNA probes for hybridisation to HG-U133 gene chip arrays. Quality control assessment of the Genechip arrays identified that scaling factors were less than 3 fold apart, ratios of 5' versus 3' probe sets for GAPDH and β-actin were close to 1 and background and noise levels were acceptable. Genes were considered to be Nurr1 dependent only if they exhibited a greater than 2 fold change (Nurr1-WT transfected cells compared to vector control) and only if their expression was attenuated by co-transfection with the Nurr1 DN expression construct (see Table 2). Only three genes were identified with this profile Interleukin-8 (IL-8), Amphiregulin (AREG) and Kit ligand (KITLG). The greatest change was seen with IL-8, which was induced 5-fold (p = 10-4).
Table 2 Differentially expressed genes following Nurr1 overexpression. K4IM cells were transfected in triplicate using the Amaxa Nucleofector system for each of the 3 conditions: 1. 5 μg pcDNA3.1 blank vector (control); 2. 2.5 μg pcDNA3.1-Nurr1-WT (Nurr1 WT) and 2.5 μg pCDNA3.1 blank vector; 3. 2.5 μg pcDNA3.1-Nurr1-DN dominant negative (DN) co-transfected with 2.5 μg pCDNA3.1-Nurr1-WT (Nurr1 DN/Nurr1 WT). Cells were cultured for 16 hours prior to RNA extraction. Genes were identified from the Affymetrix U133A chip showing significant change between blank vector transfected synoviocytes and Nurr1 transfected K4IM cells (>2 fold) and for genes not significantly changed between blank vector and Nurr1 D/N transfected cells (with a p-value of < 0.01). In each case genes were manually selected that showed subsequent return to basal levels with dominant negative cotransfection.
Official HUGO Symbol Description RefSeqN Id Fold change p-value
IL8 Interleukin 8 NM_000584 5.04 0.00047
AREG Amphiregulin (schwannoma-derived growth factor) NM_001657 2.80 0.00037
KITLG KIT ligand NM_000899 2.23 0.00390
Nurr1 dependent regulation of pro-inflammatory genes
In order to confirm the observations from the microarray analysis, K4IM cells were transfected with a Nurr1 expression construct and RNA harvested from cells after 16 hours. Quantitative RT-PCR analysis was then carried out with expression changes normalised to GAPDH. K4IM cells overexpressing Nurr1 showed increase expression of IL-8, KITLG and AREG transcripts compared to blank vector transfected cells confirming the observations from the microarray experiments (Figure 2). In addition consistent with the microarray data, the induced levels for each gene were returned to basal levels by co-transfection with the dominant negative Nurr1 construct (pcDNA3.1-Nurr1-DN) (Figure 2).
Figure 2 Effect of dominant negative Nurr1 on Nurr1 induced inflammatory gene expression. Demonstration that Nurr1 causes increased expression of IL-8, AREG and KITLG mRNA that can be attenuated with the coexpression of Nurr1-D/N in K4IM cells using Taqman Quantitative RT-PCR. Values expressed represent the mean fold change ± SEM, compared to the blank vector control and is derived from three experiments normalised in each case to GAPDH gene expression at 24 hr post transfection. Similar results were observed in a further two independent experiments.
Nurr1 dependent induction of IL-8 production in K4IM cells
To further confirm the functional consequences of elevated levels of Nurr1 we determined effects on IL-8 protein production using a sandwich ELISA on the media taken from Nurr1 transfected K4IM cells. The experiments were carried out using increasing concentrations of Nurr1 plasmid DNA (pcDNA3.1-Nurr1-WT) and a dose dependent increase in the amount of secreted IL-8 was observed (Figure 3). These results indicate that in K4IM cells elevated levels of Nurr1 leads directly to increased synthesis/secretion of the pro-inflammatory cytokine IL-8.
Figure 3 Effects on increased expression of Nurr1 on IL-8 release. 4 × 105 K4IM cells were transfected using the Amaxa Nucleofector with increasing amount of pCDNA3.1-Nurr1-WT, total amount of transfected DNA was 2 μg. Cell media was removed after 48 hour incubation and analysed for the amount of IL-8 present in the culture media using ELISA (R&D systems). Increased production of IL-8 protein secreted into the cell media was observed in a dose dependent manner with Nurr1 expression plasmid. Values expressed represent the mean concentration of IL-8 ± SEM.
Discussion
This study was designed to explore the potential link between Nurr1, pro-inflammatory signalling and the pathogenesis of rheumatoid arthritis using a synovial fibroblast derived cell line, K4IM as a model system. Transcript profiling via Affymetrix microarrays was used to examine the consequences of elevated Nurr1 levels in K4IM cells. Overexpression of Nurr1 in K4IM cells resulted in constitutive activity of the receptor as shown by the reporter assay and this activity could be suppressed by cotransfection with a dominant negative Nurr1 construct (Figure 1). Given that Nurr1 is an immediate early gene [19] and that transactivation activity using the reporter system is observed within hours of transfection (data not shown), we harvested RNA 16 hours following transfection to maximise the chance of identifying Nurr1 primary targets as opposed to downstream secondary effects. To have confidence in the Nurr1 dependent transcription of elevated genes we compared expression profiles between cells transfected with Nurr1 alone or with a mix of Nurr1 and dominant negative Nurr1. Using our stringent cutoffs, only three genes: IL-8, Amphiregulin and Kit ligand were upregulated more than 2-fold and subsequently returned to basal levels by the presence of the Nurr1 dominant negative, confirming the Nurr1 dependence. A list of probesets of 1.5-fold increase between Nurr1-WT and blank vector control transfections is provided as a supplementary table to show genes that may be either increasing or decreasing in their expression following Nurr1 transfection (see Additional file 1). This small number of differentially expressed genes may be representative of both the short time period of expression and the specificity of Nurr1 signalling. Importantly all three have recognized roles in inflammation. We examined in further detail the Nurr1-dependent induction of these genes in K4IM cells and confirmed the microarray observations using Taqman quantitative RT-PCR. Of these three genes, the most highly induced gene was the inflammatory cytokine, IL-8 and for this reason we further demonstrated using an IL-8 ELISA that Nurr1 specifically induces release of IL-8 protein into the culture media from K4IM cells (Figure 3). IL-8 has an established role in neutrophil recruitment via the CXCR1/2 receptors [29]
Amphiregulin and Kit ligand also have demonstrated roles in inflammation [30,31]. Transgenic mice expressing Amphiregulin under the control of keratin 14 promoter display early-onset synovial inflammation and severe skin pathology demonstrating a potential role for Amphiregulin in psoriasis and psoriatic arthritis [30]. Kit ligand, also known as Stem Cell Factor (SCF), has been shown to play a role in activation of mast cells. Administration of SCF into the airways of normal mice results in a dose dependent increase in airway hyperreactivity via mast cell activation demonstrating the role of SCF in the development of allergic airway inflammation and hyperreactivity [31]. In addition, activation of the Kit receptor by SCF leads to the phosphorylation of Akt which is necessary for IL-1-dependent NF-κB transactivation [32], Akt has been postulated to play a role in RA via its ability to regulate NF-κB and also promote resistance to apoptosis through a number of mechanisms [reviewed in [33]]. This role in cell cycle control remains consistent with other NR4 group members, in particular the Nur77 having an established role in TCR-mediated apoptosis of T hybridoma cells [34,35].
Nurr1 has previously been shown to act as a point of convergence for multiple inflammatory signals via CREB-1 and NFκB signalling [20]. Subsequent studies have shown Nurr1 and other NR4 family members to be involved in the inflammatory cascade of several stimuli, including TNFα-induced PAI-1 expression [36] and in LPS/TNFα stimulated macrophages [37]. TNFα plays a critical role in the stimulation of leukocyte recruitment and cytokine production and antibodies which act by blocking TNFα signalling have been shown to have clinically beneficial effects in RA patients [38]. Therefore we speculate that in addition to established NFκB signalling activating IL-8 expression [39], TNFα and other inflammation stimulators can act by regulating Nurr1 expression that in turn can regulate IL-8 expression either directly or indirectly. Ongoing work within our laboratory aims to address the exact mechanisms for Nurr1 activity on IL-8, Amphiregulin and Kit ligand gene expression.
In summary we have demonstrated using microarray analysis that elevated levels of Nurr1 leads to increased gene expression of IL-8, Amphiregulin and Kit ligand in the model cell line, K4IM. Moreover we have confirmed these Nurr1 dependent transcriptional changes using Taqman RT-PCR and demonstrated an increase in synthesis/secretion of IL-8 in cells transfected with Nurr1. We speculate that the elevated levels of Nurr1 observed in rheumatoid arthritis can potentially exacerbate the disease process in RA. Therefore, blocking the activation of Nurr1, or modifying the transactivation potential of Nurr1 through the use of rexinoids, methotrexate [40] or thiopurine analogues [41] represent a potential therapeutic option for rheumatoid arthritis and other inflammatory or allergic diseases.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MRD designed the study, carried out the experiments, analysed the data, and drafted the manuscript. CJH carried out the plasmid construction work. SR and KT carried out the microarray analysis. MDJ, AEP and MRCN participated in study design and coordination as well as editing of the manuscript. All authors have read and approved the final manuscript.
Supplementary Material
Additional file 1
Differentially expressed genes following Nurr1 overexpression. K4IM cells were transfected in triplicate using the Amaxa Nucleofector system for each of the 2 conditions: 1. 5 μg pcDNA3.1 blank vector (control); 2. 2.5 μg pcDNA3.1-Nurr1-WT (Nurr1 WT). Cells were cultured for 16 hours prior to RNA extraction. Genes were identified from the Affymetrix U133A chip showing significant change (>1.5 fold) between blank vector transfected synoviocytes and Nurr1 transfected K4IM cells (with a p-value of < 0.01). Fold changes in red are upregulated genes, those highlighted in green are downregulated genes, the previously identified genes: IL-8, AREG and KITLG are highlighted in boldface.
Click here for file
Acknowledgements
The authors wish to thank Dr. H. Eibel (U. of Freiburg) for permission to use the K4IM cells, Dr. E. Murphy (U. of Dublin) for donation of the K4IM cells and for useful discussions, to members of the RIRA arthritis exploratory team and especially to Dr. J. Wardale for critical reading of this manuscript.
==== Refs
Robinson-Rechavi M Carpentier A Duffraisse M Laudet V How many nuclear hormone receptors are there in the human genome? Trends Genet 2001 17 554 6 11585645 10.1016/S0168-9525(01)02417-9
Evans RM The steroid and thyroid hormone receptor superfamily Science 1988 240 889 95 3283939
A unified nomenclature system for the nuclear receptor superfamily Cell 1999 97 161 3 10219237 10.1016/S0092-8674(00)80726-6
Sakurada K Ohshima-Sakurada M Palmer TD Gage FH Nurr1, an orphan nuclear receptor, is a transcriptional activator of endogenous tyrosine hydroxylase in neural progenitor cells derived from the adult brain Development 1999 126 4017 4026 10457011
Perlmann T Jansson L A novel pathway for vitamin A signaling mediated by RXR heterodimerization with NGFI-B and NURR1 Genes Dev 1995 9 769 782 7705655
Maira M Martens C Philips A Drouin J Heterodimerization between members of the Nur subfamily of orphan nuclear receptors as a novel mechanism for gene activation Mol Cell Biol 1999 19 7549 57 10523643
Wang Z Benoit G Liu J Prasad S Aarnisalo P Liu X Xu H Walker NP Perlmann T Structure and function of Nurr1 identifies a class of ligand-independent nuclear receptors Nature 2003 423 555 560 12774125 10.1038/nature01645
Le WD Xu P Jankovic J Jiang H Appel SH Smith RG Vassilatis DK Mutations in NR4A2 associated with familial Parkinson disease Nat Genet 2003 33 85 89 12496759 10.1038/ng1066
Xu PY Liang R Jankovic J Hunter C Zeng YX Ashizawa T Lai D Le WD Association of homozygous 7048G7049 variant in the intron six of Nurr1 gene with Parkinson's disease Neurology 2002 58 881 884 11914402
Buervenich S Carmine A Arvidsson M Xiang F Zhang Z Sydow O Jonsson EG Sedvall GC Leonard S Ross RG Freedman R Chowdari KV Nimgaonkar VL Perlmann T Anvret M Olson L NURR1 mutations in cases of schizophrenia and manic-depressive disorder Am J Med Genet 2000 96 808 813 11121187 10.1002/1096-8628(20001204)96:6<808::AID-AJMG23>3.0.CO;2-E
Ishiguro H Okubo Y Ohtsuki T Yamakawa-Kobayashi K Arinami T Mutation analysis of the retinoid X receptor beta, nuclear-related receptor 1, and peroxisome proliferator-activated receptor alpha genes in schizophrenia and alcohol dependence: possible haplotype association of nuclear-related receptor 1 gene to alcohol dependence Am J Med Genet 2002 114 15 23 11840500 10.1002/ajmg.1620
Murphy EP McEvoy A Conneely OM Bresnihan B Fitzgerald O Involvement of the nuclear orphan receptor NURR1 in the regulation of corticotropin-releasing hormone expression and actions in human inflammatory arthritis Arthritis Rheum 2001 44 782 793 11315917 10.1002/1529-0131(200104)44:4<782::AID-ANR134>3.0.CO;2-H
Zetterstrom RH Williams R Perlmann T Olson L Cellular expression of the immediate early transcription factors Nurr1 and NGFI-B suggests a gene regulatory role in several brain regions including the nigrostriatal dopamine system Brain Res Mol Brain Res 1996 41 111 120 8883941 10.1016/0169-328X(96)00074-5
Law SW Conneely OM DeMayo FJ O'Malley BW Identification of a new brain-specific transcription factor, NURR1 Mol Endocrinol 1992 6 2129 2135 1491694 10.1210/me.6.12.2129
Wallen AA Castro DS Zetterstrom RH Karlen M Olson L Ericson J Perlmann T Orphan nuclear receptor Nurr1 is essential for Ret expression in midbrain dopamine neurons and in the brain stem Mol Cell Neurosci 2001 18 649 663 11749040 10.1006/mcne.2001.1057
Sacchetti P Dwornik H Formstecher P Rachez C Lefebvre P Requirements for heterodimerization between the orphan nuclear receptor Nurr1 and retinoid X receptors J Biol Chem 2002 277 35088 35096 12122012 10.1074/jbc.M205816200
Philips A Lesage S Gingras R Maira MH Gauthier Y Hugo P Drouin J Novel dimeric Nur77 signaling mechanism in endocrine and lymphoid cells Mol Cell Biol 1997 17 5946 5951 9315652
Kovalovsky D Refojo D Liberman AC Hochbaum D Pereda MP Coso OA Stalla GK Holsboer F Arzt E Activation and induction of NUR77/NURR1 in corticotrophs by CRH/cAMP: involvement of calcium, protein kinase A, and MAPK pathways Mol Endocrinol 2002 16 1638 1651 12089357 10.1210/me.16.7.1638
Murphy EP Conneely OM Neuroendocrine regulation of the hypothalamic pituitary adrenal axis by the nurr1/nur77 subfamily of nuclear receptors Mol Endocrinol 1997 11 39 47 8994186 10.1210/me.11.1.39
McEvoy AN Murphy EA Ponnio T Conneely OM Bresnihan B Fitzgerald O Murphy EP Activation of nuclear orphan receptor NURR1 transcription by NF-kappa B and cyclic adenosine 5'-monophosphate response element-binding protein in rheumatoid arthritis synovial tissue J Immunol 2002 168 2979 2987 11884470
Dixon G Scanlon D Cooper S Broad P A reporter gene assay for fungal sterol biosynthesis inhibitors J Steroid Biochem Mol Biol 1997 62 165 171 9393951 10.1016/S0960-0760(97)00032-0
Cerillo G Rees A Manchanda N Reilly C Brogan I White A Needham M The oestrogen receptor regulates NFkappaB and AP-1 activity in a cell-specific manner J Steroid Biochem Mol Biol 1998 67 79 88 9877207 10.1016/S0960-0760(98)00078-8
Wigler M Sweet R Sim GK Wold B Pellicer A Lacy E Maniatis T Silverstein S Axel R Transformation of mammalian cells with genes from procaryotes and eucaryotes Cell 1979 16 777 785 222468 10.1016/0092-8674(79)90093-X
Paulsen RE Weaver CA Fahrner TJ Milbrandt J Domains regulating transcriptional activity of the inducible orphan receptor NGFI-B J Biol Chem 1992 267 16491 16496 1644831
Davis IJ Hazel TG Lau LF Transcriptional activation by Nur77, a growth factor-inducible member of the steroid hormone receptor superfamily Mol Endocrinol 1991 5 854 859 1922099
Lyon JJ Watson RJ Interference of Myb transactivation activity by a conditional dominant negative protein: functional interference in a cytotoxic T-cell line results in G1 arrest Gene 1996 182 123 128 8982077 10.1016/S0378-1119(96)00531-8
Castro DS Hermanson E Joseph B Wallen A Aarnisalo P Heller A Perlmann T Induction of cell cycle arrest and morphological differentiation by Nurr1 and retinoids in dopamine MN9D cells J Biol Chem 2001 276 43277 43284 11553630 10.1074/jbc.M107013200
Haas C Aicher WK Dinkel A Peter HH Eibel H Characterization of SV40T antigen immortalized human synovial fibroblasts: maintained expression patterns of EGR-1, HLA-DR and some surface receptors Rheumatol Int 1997 16 241 247 9106935 10.1007/BF01375656
Rampart M Van Damme J Zonnekeyn L Herman AG Granulocyte chemotactic protein/interleukin-8 induces plasma leakage and neutrophil accumulation in rabbit skin Am J Pathol 1989 135 21 25 2672824
Cook PW Brown JR Cornell KA Pittelkow MR Suprabasal expression of human amphiregulin in the epidermis of transgenic mice induces a severe, early-onset, psoriasis-like skin pathology: expression of amphiregulin in the basal epidermis is also associated with synovitis Exp Dermatol 2004 13 347 356 15186320 10.1111/j.0906-6705.2004.00183.x
Campbell E Hogaboam C Lincoln P Lukacs NW Stem cell factor-induced airway hyperreactivity in allergic and normal mice Am J Pathol 1999 154 1259 1265 10233863
Cenni V Sirri A De Pol A Maraldi NM Marmiroli S Interleukin-1-receptor-associated kinase 2 (IRAK2)-mediated interleukin-1-dependent nuclear factor kappaB transactivation in Saos2 cells requires the Akt/protein kinase B kinase Biochem J 2003 376 303 311 12906710 10.1042/BJ20030028
Liu H Pope RM The role of apoptosis in rheumatoid arthritis Curr Opin Pharmacol 2003 3 317 322 12810199 10.1016/S1471-4892(03)00037-7
Liu ZG Smith SW McLaughlin KA Schwartz LM Osborne BA Apoptotic signals delivered through the T-cell receptor of a T-cell hybrid require the immediate-early gene nur77 Nature 1994 367 281 284 8121494 10.1038/367281a0
Woronicz JD Calnan B Ngo V Winoto A Requirement for the orphan steroid receptor Nur77 in apoptosis of T-cell hybridomas Nature 1994 367 277 281 8121493 10.1038/367277a0
Gruber F Hufnagl P Hofer-Warbinek R Schmid JA Breuss JM Huber-Beckmann R Lucerna M Papac N Harant H Lindley I de Martin R Binder BR Direct binding of Nur77/NAK-1 to the plasminogen activator inhibitor 1 (PAI-1) promoter regulates TNF alpha -induced PAI-1 expression Blood 2003 101 3042 3048 12506026 10.1182/blood-2002-07-2331
Pei L Castrillo A Chen M Hoffmann A Tontonoz P Induction of NR4A orphan nuclear receptor expression in macrophages in response to inflammatory stimuli J Biol Chem 2005 280 29256 29262 15964844 10.1074/jbc.M502606200
Taylor PC Peters AM Paleolog E Chapman PT Elliott MJ McCloskey R Feldmann M Maini RN Reduction of chemokine levels and leukocyte traffic to joints by tumor necrosis factor alpha blockade in patients with rheumatoid arthritis Arthritis Rheum 2000 43 38 47 10643698 10.1002/1529-0131(200001)43:1<38::AID-ANR6>3.0.CO;2-L
Georganas C Liu H Perlman H Hoffmann A Thimmapaya B Pope RM Regulation of IL-6 and IL-8 expression in rheumatoid arthritis synovial fibroblasts: the dominant role for NF-kappa B but not C/EBP beta or c-Jun J Immunol 2000 165 7199 7206 11120852
Ralph JA McEvoy AN Kane D Bresnihan B Fitzgerald O Murphy EP Modulation of orphan nuclear receptor NURR1 expression by methotrexate in human inflammatory joint disease involves adenosine A2A receptor-mediated responses J Immunol 2005 175 555 565 15972692
Wansa KD Muscat GE TRAP220 is modulated by the antineoplastic agent 6-Mercaptopurine, and mediates the activation of the NR4A subgroup of nuclear receptors J Mol Endocrinol 2005 34 835 848 15956351 10.1677/jme.1.01739
|
16309552
|
PMC1308852
|
CC BY
|
2021-01-04 16:36:23
|
no
|
J Inflamm (Lond). 2005 Nov 25; 2:15
|
utf-8
|
J Inflamm (Lond)
| 2,005 |
10.1186/1476-9255-2-15
|
oa_comm
|
==== Front
J NeuroinflammationJournal of Neuroinflammation1742-2094BioMed Central London 1742-2094-2-271630068110.1186/1742-2094-2-27Short ReportMaximal COX-2 and ppRb expression in neurons occurs during early Braak stages prior to the maximal activation of astrocytes and microglia in Alzheimer's disease Hoozemans Jeroen JM [email protected] Haastert Elise S [email protected] Robert [email protected] Thomas [email protected] Wiep [email protected] Piet [email protected] Annemieke JM [email protected] Department of Neuropathology, Academic Medical Center, P.O. Box 22700, 1100 DE Amsterdam, The Netherlands2 Neurogenetics Laboratory, Academic Medical Center, University of Amsterdam, Amsterdam, The Netherlands3 Department of Psychiatry, VU University medical center, Amsterdam, The Netherlands4 Department of Clinical Chemistry and Alzheimer Center, VU University medical center, Amsterdam, The Netherlands5 Department of Neuroanatomy, Paul Flechsig Institute for Brain Research, University of Leipzig, Leipzig, Germany2005 21 11 2005 2 27 27 24 10 2005 21 11 2005 Copyright © 2005 Hoozemans et al; licensee BioMed Central Ltd.2005Hoozemans et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Neuronal expression of cyclooxygenase-2 (COX-2) and cell cycle proteins is suggested to contribute to neurodegeneration during Alzheimer's disease (AD). The stimulus that induces COX-2 and cell cycle protein expression in AD is still elusive. Activated glia cells are shown to secrete substances that can induce expression of COX-2 and cell cycle proteins in vitro. Using post mortem brain tissue we have investigated whether activation of microglia and astrocytes in AD brain can be correlated with the expression of COX-2 and phosphorylated retinoblastoma protein (ppRb). The highest levels of neuronal COX-2 and ppRb immunoreactivity are observed in the first stages of AD pathology (Braak 0–II, Braak A). No significant difference in COX-2 or ppRb neuronal immunoreactivity is observed between Braak stage 0 and later Braak stages for neurofibrillary changes or amyloid plaques. The mean number of COX-2 or ppRb immunoreactive neurons is significantly decreased in Braak stage C compared to Braak stage A for amyloid deposits. Immunoreactivity for glial markers KP1, CR3/43 and GFAP appears in the later Braak stages and is significantly increased in Braak stage V-VI compared to Braak stage 0 for neurofibrillary changes. In addition, a significant negative correlation is observed between the presence of KP1, CR3/43 and GFAP immunoreactivity and the presence of neuronal immunoreactivity for COX-2 and ppRb. These data show that maximal COX-2 and ppRb immunoreactivity in neurons occurs during early Braak stages prior to the maximal activation of astrocytes and microglia. In contrast to in vitro studies, post mortem data do not support a causal relation between the activation of microglia and astrocytes and the expression of neuronal COX-2 and ppRb in the pathological cascade of AD.
Alzheimer's diseaseastrocytescell cyclecyclooxygenase-2microgliaretinoblastoma protein
==== Body
Findings
Aberrant expression of cyclins, cyclin dependent kinases (CDKs) and their inhibitors has been observed in post mitotic neurons in Alzheimer's disease (AD) [1,2]. Proteins that normally function to control cell cycle progression in actively dividing cells may play a role in the death of post mitotic neurons in AD [3]. The retinoblastoma protein (pRb) regulates cell proliferation by controlling progression through the restriction point within the G1-phase of the cell cycle [4]. pRb sequesters members of the E2F gene family of transcription factors. Cell cycle-dependent phosphorylation of pRb by CDKs inactivates pRb and inhibits pRb target binding, allowing cell cycle progression. The expression of phosphorylated pRb (ppRb) immunoreactivity in AD neurons has previously been described [5,6]. In the midfrontal and temporal cortex ppRb immunoreactivity can be most prominently detected in the nucleus of the large pyramidal neurons of layers III and V, and is rarely detected in neurofibrillary tangles. Recent studies have shown that neuronal cyclooxygenase-2 (COX-2) expression in AD parallels the expression of cell cycle proteins in neurons [6-8]. Previously, we observed colocalization of COX-2 with ppRb in neurons in the temporal cortex of AD and control cases [6]. Increased neuronal COX-2 expression leads to increased expression of cell cycle mediators in post mitotic neurons, as shown using a transgenic mouse model with increased neuronal COX-2 expression [9].
Once activated, microglia and astrocytes are capable of producing a variety of pro-inflammatory mediators and potentially neurotoxic substances [10], of which some have been shown to potentially induce COX-2 and cell cycle protein expression in vitro [3,11-13]. It has been shown that interleukin-1β induces COX-2 expression in neuronal cell models [11,12]. and conditioned medium from β amyloid (Aβ) peptide stimulated microglia induces expression of cell cycle proteins in neurons followed by cell death [13]. These in vitro findings indicate that the activation of microglia may play an important role in the expression of COX-2 and cell cycle proteins in neurons. Post mortem as well as in vivo studies indicate that microglial activation already occurs at an early stage in AD pathology [14,15]. Cell cycle changes and increased neuronal COX-2 expression have also been shown to be early events in AD [1,7,16,17]. We therefore hypothesized that neuronal expression of COX-2 and ppRb would be associated with increased presence and activation of glial cells.
Using post mortem brain tissue we have investigated whether activation/occurrence of microglia and astrocytes in AD brain can be correlated with the neuronal expression of COX-2 and ppRb during AD pathogenesis. Staging of AD was neuropathologicallly evaluated according to Braak and Braak [18]. Demographic characteristics of the cases used in this study are shown in table 1. For each case the area density of the immunoreactivity for KP1, CR3/43 and GFAP in the mid-temporal cortex was determined. KP1 (anti-CD68) is a marker for phagocytic microglia (and macrophages) and CR3/43 detects the class II antigens HLA-DP, DQ, DR and is generally used as a marker for activated microglia. GFAP (Glial Fibrillary Acidic Protein) is strongly and specifically expressed in astrocytes. Group summaries are expressed as box-plots for each Braak stage for neurofibrillary changes or amyloid deposits [18] (figure 1). All three markers show a gradual increase with increasing pathology. Correlation analysis reveals a statistically significant (p < 0.05) positive correlation between the Braak scores for neurofibrillary changes (NF) or Aβ deposits (AMY) and immunoreactivity for KP1 (NF, 0.671; AMY, 0.432), CR3/43 (NF, 0.564; AMY, 0.323), and GFAP (NF, 0.690; AMY, 0.424). A statistically significant increase was observed in Braak stage V-VI for KP1 (p = 0.001), CR3/43 (p = 0.008), and GFAP (p < 0.001) compared to Braak stage 0. Neuronal ppRb and COX-2 immunoreactivity are expressed as number of immunoreactive neurons per 2 mm2 (figure 1). A significant (p < 0.05) negative correlation was observed between the Braak score for neurofibrillary changes and ppRb (-0.414) or COX-2 (-0.346), and between the Braak score for Aβ plaques and COX-2 (-0.537).
Table 1 Demographic characteristics of the cases used in this study. Shown are differences between groups of the cases used in this study. [PMI post-mortem interval, SD standard deviation].
Braak score for neurofibrillary changes
O I–II III–IV V–VI
n 5 16 10 9
male/female 3/2 6/10 0/10 3/6
mean age ± SD (years) 62 ± 10 83 ± 8 89 ± 4 76 ± 7
PMI ± SD (hrs:min) 8:00 ± 4:30 7:30 ± 2:30 6:30 ± 2:30 5:00 ± 1:30
Braak score for amyloid deposits
O A B C total
n 7 6 11 16 40
male/female 4/3 3/3 3/8 2/14 12/28
mean age ± SD (years) 69 ± 12 79 ± 4 85 ± 10 82 ± 10 80 ± 11
PMI ± SD (hrs:min) 7:00 ± 4:00 8:30 ± 3:00 7:00 ± 2:30 6:00 ± 2:00 6:30 ± 2:30
Figure 1 Immunoreactivity scores for KP1, CR3/43, GFAP, ppRb and COX-2 in the temporal cortex of nondemented control and AD cases. Immunohistochemical stainings were performed as described previously [6]. The following primary antibodies were used: rabbit polyclonal anti-COX-2 (Cayman, Ann Arbor, MI), rabbit anti-phosphoserine pRb (pSer 795, Cell Signaling, Beverly, MA). Mouse anti-CD68 (KP1) and mouse anti-HLA-DP, DQ, DR (CR3/43) were obtained from DAKO (Heverlee, Belgium). Mouse anti-Glial Fibrillary Acidic Protein (GFAP) was obtained from Monosan (clone 6F2, Uden, The Netherlands). Morphometric investigation was aimed at determining the area density occupied by the immunoreactive glial cells in the cortical layer. The area density (%) was quantified using Image-Pro Plus analysis software (Media Cybernetics, Silver Spring, MD). Immunoreactive neurons (COX-2 and ppRb) were counted in a total area of 2 mm2. Neurons were distinguished from non-neuronal cells by nuclear size and shape. Values of cases are grouped according to the Braak stage for neurofibrillary changes (O, I-II, III-IV, V-VI) or Aβ deposits (O, A, B, C). Results are expressed as box plots. The box represents the interquartile range that contains 50% of the values. The whiskers extend from the box to the highest and lowest values. The line across the box indicates the median. Kruskall-Wallis test was used to evaluate differences between groups followed by the Mann-Whitney U test, to test differences between pairs of groups. Correlation analysis was done using the Pearson parametric and Spearman non-parametric method. * p < 0.05 versus Braak stage O. # p < 0.05 versus Braak stage C.
Although it is tempting to assume that these stages reflect the clinical changes, this study aims to show the relation between different molecular pathologically defined events. Cases with Braak stage A used in this study had either Braak stage I or II for neurofibrillary changes. In Braak stage A for amyloid low densities of amyloid plaques are only found in the temporal cortex and other parts of the isocortex [18]. Activated glial cells are mostly associated with neuritic plaques not with diffuse Aβ plaques [10]. This is in agreement with our data which shows a gradual increase in microglia and astrocytes with the Braak score for neurofibrillary changes and high levels of activated glial cells in cases with Braak score B and C (figure 1).
We observed maximal neuronal ppRb and COX-2 immunoreactivity in Braak stages 0 and A. No significant difference in ppRb and COX-2 immunoreactivity was observed between the Braak stages for neurofibrillary changes. The maximal ppRb and COX-2 immunoreactivity in stage A did not significantly differ from stage O. However, we did observe a significant decrease in Braak stage C compared to stage A. These findings contradict previous studies that have shown increased neuronal COX-2 expression [19,20] and ppRb immunoreactivity in AD cases [5]. In the present study the patients are grouped according to the Braak stage instead of being defined as control or AD. Other, previously described [17], discrepancies are most likely due to differences in pathological disease state and investigated brain area, methods of analysis, as well as technical issues. The data presented in this study are in agreement with the findings of Yermakova and O'Banion [17]. In an immunohistochemical study they found a decrease in the number of COX-2 immunoreactive neurons in advanced stages of AD. A similar trend, as shown in the present study, was observed in the hippocampus comparing the mean neuronal COX-2 immunoreactivity with the Braak score for NF. A non-significant higher mean level in Braak stage I-II was also reported [17]. The levels of neuronal COX-2 expression observed in post mortem brain tissue correlate well with recent clinical data presented by Combrinck and colleagues [21] describing, compared to control patients, higher prostaglandin E2 levels in the cerebrospinal fluid in patients with mild memory impairment, but lower in those with more advanced AD.
A significant negative correlation was observed between the area density of KP1 and the immunoreactivity for ppRb (-0.414, p = 0.007) and COX-2 (-0.366, p = 0.020). These data suggest no (positive) relation between neuronal expression of COX-2 or ppRb and the increased glial response observed during AD pathology. Although suggested by in vitro studies, our evaluation of post mortem brain tissue suggests that it is very unlikely that activation of microglia or astrocytes cause neuronal expression of COX-2 and ppRb in AD. Although the involvement of activated glia in the initial upregulation of these factors seems unlikely, we cannot exclude the involvement of glia in the regulation of COX-2 or cell cycle protein expression in neurons at later stages of pathology.
COX-2 and cell cycle changes can be detected in neurons that are vulnerable for developing neurodegenerative changes that are associated with AD [6,16,22]. This implies that COX-2 and neuronal cell cycle changes occur in the early steps of AD neurodegeneration. Moreover, high levels of neuronal COX-2, ppRb, cyclin D1 and cyclin E are found in the temporal cortex of cases which have diffuse Aβ deposits while fibrillar/neuritic plaques are absent [6,7]. Various in vitro studies using neuronal models show that Aβ peptide induces COX-2 [20] and phosphorylation of pRb [23,24], which is followed by neuronal cell death. In this perspective, the current emerging data on the early role of oligomeric and protofibrilic forms of Aβ in AD is very interesting [25,26]. Whether COX-2 and cell cycle proteins are part of the molecular mechanisms involved in the response to intraneuronal accumulation of Aβ and the consequent impaired synaptic function needs to be addressed in future studies.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
JJMH participated in the design of the study, performed the statistical analysis and prepared the manuscript. ESvH carried out the immunohistochemical analyis and quantification of the immunohistochemical data. RV has been involved in the collection of the human post mortem brain material. TA, WS, PE and AJMR participated in the design of the study and helped to draft the manuscript. All authors read and approved the final manuscript.
Acknowledgements
The authors thank the Netherlands Brain Bank for supplying the human brain tissue (coordinator Dr. R. Ravid) and Dr. W. Kamphorst for the neuropathological diagnosis of control and AD tissue. This study was supported by the Internationale Stichting Alzheimer Onderzoek (ISAO grant 04503 to JJMH) and the European Community (ADIT programme, LSHB-CT-511977 to AJMR).
==== Refs
Nagy Z Esiri MM Cato AM Smith AD Cell cycle markers in the hippocampus in Alzheimer's disease Acta Neuropathol (Berl) 1997 94 6 15 9224524 10.1007/s004010050665
Arendt T Rodel L Gartner U Holzer M Expression of the cyclin-dependent kinase inhibitor p16 in Alzheimer's disease Neuroreport 1996 7 3047 3049 9116237
Arendt T Synaptic plasticity and cell cycle activation in neurons are alternative effector pathways: the 'Dr. Jekyll and Mr. Hyde concept' of Alzheimer's disease or the yin and yang of neuroplasticity Prog Neurobiol 2003 71 83 248 14687983 10.1016/j.pneurobio.2003.09.007
Sherr CJ Cancer cell cycles Science 1996 274 1672 1677 8939849 10.1126/science.274.5293.1672
Jordan-Sciutto KL Malaiyandi LM Bowser R Altered distribution of cell cycle transcriptional regulators during Alzheimer disease J Neuropathol Exp Neurol 2002 61 358 367 11939591
Hoozemans JJ Veerhuis R Rozemuller AJ Arendt T Eikelenboom P Neuronal COX-2 expression and phosphorylation of pRb precede p38 MAPK activation and neurofibrillary changes in AD temporal cortex Neurobiol Dis 2004 15 492 499 15056456 10.1016/j.nbd.2003.11.028
Hoozemans JJ Bruckner MK Rozemuller AJ Veerhuis R Eikelenboom P Arendt T Cyclin D1 and cyclin E are co-localized with cyclo-oxygenase 2 (COX-2) in pyramidal neurons in Alzheimer disease temporal cortex J Neuropathol Exp Neurol 2002 61 678 688 12152783
Mirjany M Ho L Pasinetti GM Role of cyclooxygenase-2 in neuronal cell cycle activity and glutamate-mediated excitotoxicity J Pharmacol Exp Ther 2002 301 494 500 11961048 10.1124/jpet.301.2.494
Xiang Z Ho L Valdellon J Borchelt D Kelley K Spielman L Aisen PS Pasinetti GM Cyclooxygenase (COX)-2 and cell cycle activity in a transgenic mouse model of Alzheimer's disease neuropathology Neurobiol Aging 2002 23 327 334 11959394 10.1016/S0197-4580(01)00282-2
Akiyama H Barger S Barnum S Bradt B Bauer J Cole GM Cooper NR Eikelenboom P Emmerling M Fiebich BL Finch CE Frautschy S Griffin WS Hampel H Hull M Landreth G Lue L Mrak R Mackenzie IR McGeer PL O'Banion MK Pachter J Pasinetti G Plata-Salaman C Rogers J Rydel R Shen Y Streit W Strohmeyer R Tooyoma I Van Muiswinkel FL Veerhuis R Walker D Webster S Wegrzyniak B Wenk G Wyss-Coray T Inflammation and Alzheimer's disease Neurobiol Aging 2000 21 383 421 10858586 10.1016/S0197-4580(00)00124-X
Fiebich BL Mueksch B Boehringer M Hull M Interleukin-1beta induces cyclooxygenase-2 and prostaglandin E(2) synthesis in human neuroblastoma cells: involvement of p38 mitogen-activated protein kinase and nuclear factor-kappaB J Neurochem 2000 75 2020 2028 11032891 10.1046/j.1471-4159.2000.0752020.x
Hoozemans JJ Veerhuis R Janssen I Rozemuller AJ Eikelenboom P Interleukin-1beta induced cyclooxygenase 2 expression and prostaglandin E2 secretion by human neuroblastoma cells: implications for Alzheimer's disease Exp Gerontol 2001 36 559 570 11250126 10.1016/S0531-5565(00)00226-6
Wu Q Combs C Cannady SB Geldmacher DS Herrup K Beta-amyloid activated microglia induce cell cycling and cell death in cultured cortical neurons Neurobiol Aging 2000 21 797 806 11124423 10.1016/S0197-4580(00)00219-0
Arends YM Duyckaerts C Rozemuller JM Eikelenboom P Hauw JJ Microglia, amyloid and dementia in alzheimer disease. A correlative study Neurobiol Aging 2000 21 39 47 10794847 10.1016/S0197-4580(00)00094-4
Cagnin A Brooks DJ Kennedy AM Gunn RN Myers R Turkheimer FE Jones T Banati RB In-vivo measurement of activated microglia in dementia Lancet 2001 358 461 467 11513911 10.1016/S0140-6736(01)05625-2
Busser J Geldmacher DS Herrup K Ectopic cell cycle proteins predict the sites of neuronal cell death in Alzheimer's disease brain J Neurosci 1998 18 2801 2807 9525997
Yermakova AV O'Banion MK Downregulation of neuronal cyclooxygenase-2 expression in end stage Alzheimer's disease Neurobiol Aging 2001 22 823 836 11754989 10.1016/S0197-4580(01)00303-7
Braak H Braak E Neuropathological stageing of Alzheimer-related changes Acta Neuropathol (Berl) 1991 82 239 259 1759558 10.1007/BF00308809
Oka A Takashima S Induction of cyclo-oxygenase 2 in brains of patients with Down's syndrome and dementia of Alzheimer type: specific localization in affected neurones and axons Neuroreport 1997 8 1161 1164 9175105
Pasinetti GM Aisen PS Cyclooxygenase-2 expression is increased in frontal cortex of Alzheimer's disease brain Neuroscience 1998 87 319 324 9740394 10.1016/S0306-4522(98)00218-8
Combrinck M Williams J De Berardinis MA Warden D Puopolo M Smith AD Minghetti L Levels of CSF prostaglandin E2, cognitive decline and survival in Alzheimer's disease J Neurol Neurosurg Psychiatry 2005 15944180
Yang Y Mufson EJ Herrup K Neuronal cell death is preceded by cell cycle events at all stages of Alzheimer's disease J Neurosci 2003 23 2557 2563 12684440
Giovanni A Wirtz-Brugger F Keramaris E Slack R Park DS Involvement of cell cycle elements, cyclin-dependent kinases, pRb, and E2F x DP, in B-amyloid-induced neuronal death J Biol Chem 1999 274 19011 19016 10383401 10.1074/jbc.274.27.19011
Copani A Condorelli F Caruso A Vancheri C Sala A Giuffrida Stella AM Canonico PL Nicoletti F Sortino MA Mitotic signaling by beta-amyloid causes neuronal death Faseb J 1999 13 2225 2234 10593870
Klyubin I Walsh DM Lemere CA Cullen WK Shankar GM Betts V Spooner ET Jiang L Anwyl R Selkoe DJ Rowan MJ Amyloid beta protein immunotherapy neutralizes Abeta oligomers that disrupt synaptic plasticity in vivo Nat Med 2005 11 556 561 15834427 10.1038/nm1234
Gouras GK Almeida CG Takahashi RH Intraneuronal Abeta accumulation and origin of plaques in Alzheimer's disease Neurobiol Aging 2005 26 1235 1244 16023263 10.1016/j.neurobiolaging.2005.05.022
|
16300681
|
PMC1308853
|
CC BY
|
2021-01-04 16:38:21
|
no
|
J Neuroinflammation. 2005 Nov 21; 2:27
|
utf-8
|
J Neuroinflammation
| 2,005 |
10.1186/1742-2094-2-27
|
oa_comm
|
==== Front
Malar JMalaria Journal1475-2875BioMed Central London 1475-2875-4-531627115310.1186/1475-2875-4-53ResearchMalaria chemoprophylaxis and the serologic response to measles and diphtheria-tetanus-whole-cell pertussis vaccines Rosen Jennifer B [email protected] Joel G [email protected] Charles R [email protected] Bruce D [email protected] William E [email protected] Hans O [email protected] Pierre [email protected] Jacquelin M [email protected]é Pierre [email protected] Mark A [email protected] Division of International Epidemiology and Population Studies, Fogarty International Center, National Institutes of Health, Bethesda, MD 20892, USA2 Howard Hughes Medical Institute-National Institutes of Health Research Program, Bethesda, MD 20892, USA3 Division of Bacterial Products, Allergenic and Parasitic Products, Center for Biologics Evaluation and Research, Food and Drug Administration, Bethesda, MD 20892, USA4 Division of Parasitic Diseases, National Center for Infectious Diseases, Centers for Disease Control and Prevention, Atlanta, GA 30333, USA5 Société de Pathologie Exotique, Paris, France6 01 BP 1587 Ouagadougou 01, Burkina Faso2005 6 11 2005 4 53 53 14 7 2005 6 11 2005 Copyright © 2005 Rosen et al; licensee BioMed Central Ltd.2005Rosen et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Acute malaria has been associated with a decreased antibody response to tetanus and diphtheria toxoids, meningococcal, salmonella, and Hib vaccines. Interest in giving malaria drug therapy and prevention at the time of childhood immunizations has increased greatly following recent trials of intermittent preventive therapy during infancy (IPTi), stimulating this re-analysis of unpublished data. The effect of malaria chemoprophylaxis on vaccine response was studied following administration of measles vaccines and diphtheria-tetanus-whole cell pertussis (DTP) vaccines.
Methods
In 1975, six villages divided into two groups of children ≤74 months of age from Burkina Faso, were assigned to receive amodiaquine hydrochloride chemoprophylaxis (CH+) every two weeks for seven months or no chemoprophylaxis (CH-). After five months, children in each group received either one dose of measles or two doses of DTP vaccines.
Results
For recipients of the measles vaccine, the seroconversion rates in CH+ and CH- children, respectively, were 93% and 96% (P > 0.05). The seroresponse rates in CH+ and CH- children respectively, were 73% and 86% for diphtheria (P > 0.05) and 77% and 91% for tetanus toxoid (P > 0.05). In a subset analysis, in which only children who strictly adhered to chemoprophylaxis criteria were included, there were, likewise, no significant differences in seroconversion or seroresponse for measles, diphtheria, or tetanus vaccines (P > 0.05). While analysis for pertussis showed a 43% (CH+) and 67% (CH-) response (P < 0.05), analyses using logistic regression to control for sex, age, chemoprophylaxis, weight-for-height Z-score, and pre-vaccination geometric mean titer (GMT), demonstrated that chemoprophylaxis was not associated with a significantly different conversion rate following DTP and measles vaccines. Seven months of chemoprophylaxis decreased significantly the malaria IFA and ELISA GMTs in the CH+ group.
Conclusion
Malaria chemoprophylaxis prior to vaccination in malaria endemic settings did not improve or impair immunogenicity of DTP and measles vaccines. This is the first human study to look at the association between malaria chemoprophylaxis and the serologic response to whole-cell pertussis vaccine.
==== Body
Background
Malaria accounts for an estimated 1 to 3 million deaths each year, with the majority occurring in children under five years of age in sub-Saharan Africa [1]. Vaccine-preventable diseases cause an estimated 1 to 2 million deaths in African children [2]. The WHO's Expanded Program on Immunization (EPI) is targeted at malarious areas, emphasizing the need to understand the effect of malaria and antimalaria drug use on vaccine immunogenicity and efficacy. Accordingly, a study that began in 1975 has been fully analysed following great increasing recent interest in the important topic of malaria chemoprophylaxis and, in particular, intermittent preventive (malaria) therapy of infants (IPTi) [3-7].
Acute malaria has been associated with a decreased response to tetanus toxoids, and meningococcal polysaccharide, Hib conjugate, and whole cell vaccines for typhoid fever [8-10]. Asymptomatic parasitaemia has been associated with a decreased response to the newer acellular pertussis and meningococcal vaccines, suggesting a benefit from malaria prophylaxis prior to vaccination [11-13]. Other studies have shown that asymptomatic parasitaemia or anti-malarial drug administration does not inhibit vaccine response to various live, attenuated, whole-cell killed, and toxoid vaccines [4,5,14-20]. No human studies have looked at the association between parasitaemia and the serologic response to whole-cell pertussis vaccine, a product still used in many vaccination programmes, particularly in developing countries. Antimalarials may also depress vaccine response as illustrated by the immunodepressive effect of 4-aminoquinolones[13,21-24].
The study aimed to determine the effect of malaria chemoprophylaxis on vaccine seroconversion or seroresponse to live, attenuated measles vaccine, diphtheria and tetanus toxoids and whole-cell pertussis (DTP) vaccines.
Methods
Study area and population
The study was conducted from May through December in 1975 in six villages; all were located in the Guinean savanna and were hyper- and holo-endemic for malaria, depending on transmission season [25]. Before the study began (February-March, during the low transmission season), a 52% Plasmodium falciparum parasitaemia prevalence was found in 150 children (25 per site) <6 years of age, with no major differences between the sites; during this pre-study investigation, antibodies to P. falciparum were detected by indirect haemagglutination (IHA) in 100 percent of children tested from five of the six villages (25 children per village). Burkinabe clinicians in the nearest dispensaries and hospitals stated that the study area was endemic for measles (cases and deaths occurred during the study), diphtheria, tetanus, and pertussis, but the incidence was unknown; routine data had not been collected from the study villages because the EPI had not yet begun [26]. Hence, previous vaccination of children was extremely unlikely and was confirmed by interrogation of individual families. There was no malaria prophylaxis; treatment for fevers and other illness was obtained from traditional healers and in dispensaries within five km of the villages. P. falciparum resistance to 4-aminoquinolones was unknown in the area. Exclusion criteria for participation in the study included acute or chronic severe illness and the presence of detectable pre-vaccination measles antibodies. Children were weighed with a calibrated hanging scale. The length of very young children was measured with them lying down on a calibrated board and, for older children, standing, according to Centers for Disease Control and Prevention (CDC) nutrition programme guidelines.
Study design
Children, aged 4 to 74 months (N = 996) living in the 6 malarious villages were assigned to 2 groups of 3 villages each, depending on whether they lived in villages east or west of Bobo-Dioulasso. One group received amodiaquine prophylaxis, CH+ [N = 488], and the other received no prophylaxis, CH- [N = 508]. Virtually all children within the same village received either measles [N = 482], or DTP vaccine [N = 514] (Figure 1). The study villages had differing populations of target-aged children. Hence, village and subject separation into measles versus DTP groups was based on the estimated number of children needed for each vaccine, and was dependent on the initial calculation of sample size (see Statistical Analysis). There was no blinding of study participants or researchers.
Figure 1 Malaria chemoprophylaxis and response to childhood vaccinations: flow chart of study design. Number of subjects excluded from analysis due to moving, death or lack of available sera: 1. 4 (37 excluded due to detectable pre-vaccination measles antibody titers) 2. 163 3. 177 4. 162 5. 70 6. 66 7. 36 8. 31 (68 excluded due to detectable pre-vaccination measles antibody titers, 18 excluded due to receipt of some prophylaxis)
The populations were of a related Mandinke ethnic group (Bobo in the eastern and Senufo in the western villages) and the communities had similar Anopheles gambiae ecology. Villages were chosen, based on the level of endemicity of malaria, the immunologic naivete of those receiving immunizations, and their proximity to one another; villages were spaced far enough apart so that families in CH- villages did not know that children in CH+ villages were receiving chemoprophylaxis, yet were close enough (within 75 km of Bob-Dioulasso) for travel convenience and management by the research team.
Chemoprophylaxis
CH+ children received a single prophylactic dose of amodiaquine hydrochloride suspension or tablets every 2 weeks for 7 months beginning in May and June, the start of the transmission season; those <12 months of age received 100 mg, those 12 to 47 months received 200 mg, and those 48 to 73 months old received 300 mg. Both amodiaquine and chloroquine are 4-aminoquinolines. Amodiaquine is the prodrug for the active ingredient desethylamodiaquine (DAQ). DAQ has a terminal half life of one to three weeks and is schizonticidal in very low concentrations [27,28]. While amodiaquine lost favour because of its association with agranulocytosis, the drug is now being re-evaluated. Because amodiaquine has not been used for over 2 decades, it has somewhat greater efficacy than chloroquine for P. falciparum resistant to chloroquine [29,30].
Seroconversion rates were measured in all CH+ and CH- children for whom paired sera were available. In addition, a CH+ group having strict compliance to drug ingestion was analyzed to examine more carefully the effect of malaria or drug use on serconversion and seroresponse. For this study, criteria for strict compliance included: receipt of ≥75% of the doses, no two consecutive doses missing, no doses missed in the month prior to vaccination(s), and no doses missed between the second vaccination and the last blood draw. At study completion, during the beginning of the low transmission season, families of children in the CH+ group were given a short-term supply of amodiaquine prophylaxis and instructions for home treatment in case of a rebound malaria attack. Follow-up visits occurred every one to two months in all villages for six months after the study.
Vaccination
All children were vaccinated with either measles vaccine or the first dose of DTP at month 5 (October or November, peak malaria transmission) and the second DTP dose one month later. Two doses of DTP were given during the study (rather than the standard initial series of three doses) to increase the likelihood that any effect from chemoprophylaxis would be discerned; a third dose was given at the end of the study but bloods were not drawn after this third vaccination. Vaccinations and amodiaquine were administered on the same day. The manufacturer and source of the licensed vaccines used were the Dow, Lirugen (Schwarz strain) measles vaccine, Lot No. 185806 AA and Merrill-National DTP, adsorbed, USP, vaccine, filling number 1036 DM, bulk Lot No. 1832. Measles and DTP vaccines were injected intramuscularly (0.5 ml) via a hydraulic (pressurized), injection device, the Ped-O-Jet® injector. Following study completion, all children in participating villages received the vaccine that they did not receive during the study. Prior to use, the measles and DTP vaccines were tested for potency and met standard requirements.
Vaccine serology
Venous blood samples were kept refrigerated and within 1 to 3 days serum was separated and kept at -20°C prior to shipment to the CDC, Atlanta, Georgia, USA, in dry ice. Serologic testing for the measles vaccine was performed in 1977 at the then Virology Division, Bureau of Laboratories, CDC. Antibody response to the DTP vaccine was performed in 1977 at the then Division of Bacterial Products, Bureau of Biologics, Food and Drug Administration, in Bethesda, MD.
Measles antibody was assessed by haemagglutination inhibition [31-33]. Seroconversion was defined by a rise in titer to ≥1:20 from an initial titer of <1:10 (the lowest detectable titer). Diphtheria and tetanus antibody titers were measured by passive microhemagglutination, using tanned sheep red blood cells [34,35]. Individuals showing a >4-fold rise in antibodies to uncoated sheep red blood cells were not included in the analysis. Pertussis titers were measured by microagglutination using killed cells of Bordetella pertussis strains 134 and 165 [36]. The titer is the log2 of the reciprocal of the highest final serum dilution resulting in detectable agglutination. When sufficient serum was available, the lowest final serum dilution tested was 1:8; by convention, samples negative at a 1:8 dilution were assigned the titer log2 = 2. If the serum sample volume was low, higher initial dilutions were used. When such sera were negative at the lowest dilution tested, the titer was reported as <lowest dilution tested. Because the actual end-point was not known, titers for these sera were not included in the calculation of geometric mean titer (GMT) or geometric mean fold rise in titer (GMR). For some individuals, it was possible to verify a ≥4-fold rise even if the actual endpoint was not known for both sera. A ≥4-fold rise for DTP antigens was considered positive. For DTP, in cases where the titer decreased from pre- to post-vaccination, a ΔGMT = 0 was used in the calculation of GMR.
Malaria serology
P. falciparum IgG antibodies were measured by the Immunofluorescence Assay (IFA) [37] and by the Enzyme-Linked Immunosorbent Assay (ELISA) [38] following collection of blood 7 months after the children were CH+ or CH- status.
Statistical Analysis
Sample size was determined initially by a method comparing two proportions. For measles, 215 subjects were required for each group in order to have 90% assurance of significant results to detect this 10% difference in response rates. Similarly, for DTP vaccines, assuming 70% seroconversion for the test group and 60% for the control group, 387 subjects were needed for each group. SAS software, version 9.00 (SAS Institute, USA), was used. Weight-for-height Z-score was calculated using an anthro-system (version 1.02, WHO-CDC, Switzerland).
The primary outcome was rate of seroconversion or seroresponse in CH+ and CH- individuals. As secondary outcomes, geometric mean titers (GMT) and mean fold rise in titer (GMR) were measured for measles, DTP, and malaria antibodies for CH+ and CH- individuals for the strict compliance group.
Study population characteristics at vaccination were compared for the CH+ and CH- children using the Chi-squared test and Student's pooled t-test. The Chi-squared test was used for comparison of seroconversion to measles vaccine and seroresponse to DTP vaccinations; the Student's pooled t-test was used for pre- and post-vaccination GMT and GMR. A univariate logistic analysis was performed to assess effects of sex, age (> or < 24 months), prophylaxis, weight-for-height Z-score (> or < median Z-score), and pre-vaccination GMT (> or < median GMT) on seroconversion or seroresponse. Multivariate logistic regression was performed on those factors found to be independently associated with seroresponse. Analyses for seroconversion, GMTs, GMRs, and logistic regression were adjusted for village effect.
Consent
The study protocol was approved by the Burkina Faso (Upper Volta) Ministry of Health and the Institutional Review Board at the CDC. Verbal permission for the study was obtained from the village chiefs, their "council of elders," and each participating family, as was the custom for working in Burkinabe villages.
Results
The groups were similar at baseline with regard to age, sex, and nutritional status, except for a slight excess of males in the CH- group for DTP (Table 1). Twenty percent of the children were <12 months of age (N = 202). Figure 1 shows the distribution flow of children receiving or not receiving chemoprophylaxis by vaccine type. Although the vaccine was administered as combined DTP, the number of children with serological data that were evaluated differed for the three DTP assays as defined above. The final number of children analysed reflects the availability of paired sera, or loss due to moves or death (Figure 1). When comparing the compliant CH+ children to the non-compliant CH+ children, there were no significant differences in sex, age, or weight-for-height Z-score for those children receiving measles vaccine (P = 0.33, 0.56, 0.52 respectively) or in sex for children receiving DTP (P = 0.22). For the DTP group, age and weight-for-height Z-score was significantly less in the noncompliant CH+ group (P < 0.01, P = 0.02, respectively).
Table 1 Characteristics of children qualifying for analyses at vaccination by vaccine type, gender, age and nutrition (measured by the Weight-for-Height Z-score at vaccination) for the prophylaxis (CH+) and no prophylaxis (CH-) groups.
Prophylaxis (CH+) No Prophylaxis (CH-)
Vacci ne Type Trait # % or Mean (SD*) # % or Mean (SD*) P-value
Measles Sex % male 177 53 274 54 0.82
Age in months 178 33.7 (15.2) 274 33.7 (18.1) 0.97
Wt-for-Ht Z-score 162 -0.65 (0.92) 247 -0.54 (1.01) 0.28
DTP Sex % male 310 48 204 53 0.04
Age in months 309 32.3 (14.9) 204 32 (17.7) 0.29
Wt-for-Ht Z-score 210 -0.88 (0.98) 37 -0.78 (0.84) 0.29
* SD Standard Deviation
Table 2 shows that seroconversion rates to measles vaccine and seroresponse to diphtheria and tetanus vaccines were not significantly different in the CH+ and CH- groups, both when all children were included and when non-compliant CH+ children were excluded from the analysis (P > 0.05). When all children were included in this analysis, there was a lower rate of seroconversion to diphtheria and tetanus in the CH+ group, but this difference was not statistically significant. Percent seroresponse to pertussis was greater in the CH- group (P < 0.01). In this cluster analysis, there was adjustment for the random effect of village. When the analysis was done without adjusting for the random effect of village, there was a significantly greater rate of seroconversion to diphtheria and tetanus in the CH- group. When non-compliant children were excluded, the difference was no longer significant for diphtheria and tetanus. Percent seroresponse to pertussis remained greater in the CH- group (P < 0.01).
Table 2 Seroconversion to measles, diphtheria, tetanus, and pertussis vaccinations in the prophylaxis (CH+) and no prophylaxis (CH-) groups
Vaccine Proportion With Seroconversion or Seroresponse (%)† P-value
(Adjusted For Village) Relative Risk (95% CI)
CH+ CH-
Measles 127/137(93)* 180/187 (96) 0.16* 0.96 (0.91–1.02)*
109/116 (94) 0.36 0.98 (0.92–1.03)
Diphtheria 108/147(73)* 116/135 (86) 0.26* 0.86 (0.76–0.96)*‡
38/ 46 (83) 0.59 0.96 (0.83–1.12)
Tetanus 104/134(77)* 126/138 (91) 0.08* 0.85 (0.77–0.94)*‡
41/43 (95) 0.39 1.04 (0.96–1.14)
Pertussis 63/148(43) * 113/168 (67) <0.01* 0.63 (0.51–0.78)*
17/44 (39) <0.01 0.57 (0.39–0.85)
* top rows include all CH+ children (bottom rows include only those CH+ children who met criteria for strict compliance)
† seroconversion is any increase in titer from a negative baseline for measles; seroresponse is a four-fold or greater rise in titer for diphtheria, tetanus, and pertussis
‡ the discordance between the P-value and confidence interval arises because the latter was calculated from a separate analysis which did not adjust for the random effect of village
For measles, pre-vaccination measles titers for all children were <1:10 (lowest detectable titer) (Figure 2); GMR was not significantly different in the CH+ vs. CH- group (P = 0.44). For all three antigens GMR was higher in the CH- group, but this difference was statistically significant only for pertussis (P < 0.01).
Figure 2 Pre and post-vaccination Geometric Mean Titers (GMTs) and Geometric Mean Fold Rise (GMR) for prophylaxis (CH+) and no prophylaxis (CH-) groups. P values listed correspond to the difference in GMR between the two groups. GMTs expressed as log2. Children included in the CH+ group met criteria for compliance for chemoprophylaxis. (A) Measles vaccine: GMR for CH+ and CH- groups does not differ significantly. (B) Diphtheria vaccine: GMR for CH+ and CH- groups does not differ significantly. (C) Tetanus vaccine: GMR for CH+ and CH- groups does not differ significantly. (D) Pertussis vaccine: GMR is significantly > for CH- group.
Multivariate logistic regression analysis indicated that for tetanus, a lower pre-vaccination GMT was positively associated with seroresponse (P = 0.02); for pertussis, a lower pre-vaccination GMT (P < 0.01) and younger age (P = 0.04) were positively associated with seroresponse. There was no significant difference in pre-vaccination pertussis titres between the CH+ and CH- group (P 0.22) when looking at all children regardless of compliance; however, when excluding non-compliant children, pre-vaccination pertussis titres were higher in CH+ children (P <0.01). While pre-vaccination pertussis titres were higher in compliant CH+ children (log2 of GMT = 6.19, N = 37) compared to non-compliant CH+ children (log2 of GMT = 5.80, N = 84), this difference was not significantly different (P = 0.80). Chemoprophylaxis was not associated with seroresponse for any of the vaccines. Thus, especially for pertussis, the lower vaccine response rate observed in the CH+ group appears to be due, in part, to the greater proportion of subjects with high pre-immunization titers.
Malaria antibody titers were significantly lower in the CH+ group compared to the CH- group following seven months of prophylaxis. GMTs for children receiving measles vaccine were: CH+, 196 (N = 128) vs. CH-, 1089 (N = 219) (P < 0.01) using IFA and CH+, 285 (N = 60) vs CH-, 1990 (N = 64) using ELISA (P < 0.01) and for children receiving DTP vaccine: CH+, 109 (N = 132) vs CH-, 178 (N = 178) (P < 0.01) using IFA and CH+, 86 (N = 63) vs CH-,153 (N = 30) (P = 0.01) using ELISA. Only 13 percent of all CH+ children with detectable malaria titers prior to chemoprophylaxis had undetectable titers post-chemoprophylaxis (N = 159). This indicates that chemoprophylaxis given to young children for five to seven months after previous exposure to malaria was not adequate to eliminate malaria antibodies.
No adverse events were recorded after chemoprophylaxis, blood sample collection, or vaccination other than a few children with 1–3 mm nodules on their arms after receiving the vaccine by injector and one child who developed a cellulitis where the venopuncture occurred: this child recovered with systemic antibiotic treatment.
Discussion
Proposed mechanisms for malaria-associated immunodepression include impaired macrophage function [39-41], altered cytokine production [39,42], a depletion of T or B cells [43], impaired dendritic cells[42,44,45], elevated nitric oxide production [46] and elevated prostaglandin E during febrile malaria[47]. Clinical evidence includes an association of malaria with increased susceptibility to bacterial infections [48], reactivation of viral infections [49,50], a low prevalence of autoimmune disease in endemic areas[51,52], and reports of decreased responses to vaccinations.
In contrast to asymptomatic parasitaemia, acute malaria impairs vaccine response[8-10,12,17,18,20]. In vitro challenge studies in individuals with acute malaria have demonstrated a depression in the cellular immune response involving alterations in lymphoproliferation and cytokine responses [42,53,54]. The pyrogenic cytokine TNF-alpha is elevated in febrile malaria and may depress humoral immunity by impairing antigen handling by dendritic cells. T-cell levels, CD4 cells in particular, are depressed [55]. IL-1, in addition to TNF-alpha, is elevated in acute illnesses [56,57]. Both promote increased T-cell adhesion to endothelium, which may lead to T-cell margination and sequestration and, thus, a decrease of functional T-cells [55]. CD4 cells secrete cytokines that activate CD8 cells, B-cells and macrophages. In acute malaria, a depression of CD4 cells leads to depressed cellular and humoral immunity, impairing vaccine response.
The absence of association between malaria chemoprophylaxis and vaccine response in this study is consistent with findings from other chemoprophylaxis studies in malarious areas involving children with asymptomatic parasitaemia [4,5,14,16-18,58]. No prior studies have published data on the association between chemoprophylaxis and pertussis (killed, whole-cell) vaccine response in humans.
The agglutination test remains the test of choice for whole-cell pertussis vaccines. Although the antigen-specific ELISA tests can amplify the information provided by the agglutination test, the agglutination test is the only one that has been shown clinically to correlate with vaccine efficacy of whole-cell pertussis vaccines and has been used in relatively recent trials [59,60]. Although ELISA or cell-culture based methods are more widely used today than the passive haemagglutination method for tetanus and diphtheria antitoxins, passive haemagglutination remains acceptable for evaluations of immunized populations [61]. Had they been available, the newer serological tests for measles, including neutralization testing, would have provided greater insight regarding clinical protection from disease.
Results of this study indicated that for the pertussis component, children <24 months of age had a better seroresponse. Vaccinating children <24 months of age will more effectively target the population in greatest need. Pertussis is most serious for very young infants and because complications leading to hospitalization, pneumonias, and death occur most often in those <24 months of age, the recommended age for initiation of pertussis immunization is generally two to three months.
Three doses of DTP vaccine comprise the usual primary series;thus, it would have been useful to assess seroconversion after a third DTP dose in addition to the response following the second dose. Technical and logistical considerations precluded this; additionally, there was concern regarding the possibility of decreased compliance with a longer study, as well as the potential to minimize any differences in the effect of chemoprophylaxis on seroconversion.
Malaria serologies demonstrated a significant difference in GMTs between the two treatment groups at the time of vaccination, adding some assurance that chemoprophylaxis decreased asymptomatic parasitaemia. Despite assumed effective chemoprophylaxis for five to seven months, virtually all compliant children had malaria antibodies; this probably reflected a durable antibody response to infections acquired before the study began. While not the primary study objective, fever prevalence data and blood smear records would have provided valuable insight on malaria prevention in the chemoprophylaxis group. Given efforts to administer intermittent preventive therapy of infants (IPTi) in conjunction with the vaccines included in the EPI, additional prospective studies are needed to establish more firmly the effect of antimalarials on response to childhood vaccinations [6].
Conclusion
Malaria chemoprophylaxis does not appear to enhance nor impair the antibody response to DTP and measles vaccines. There have been several changes over the 30 years since the study completion, including development of falciparum malaria resistance to 4-aminoquinolones throughout Africa, and establishment of the EPI (1977) and the Roll Back Malaria Partnership (1998). The continuing development and deployment of new antimalarial drugs and childhood vaccines mandates that the possible immunologic and protective interrelationships of these new products be investigated. Studies are in progress by the IPTi Consortium to address these issues [7].
Authors' contributions
Dr. Breman was responsible for writing the protocol, carrying out the study in the field, data analysis and writing; Dr. Rosen for synthesis, data analysis and writing; Dr. Manclark for coordinating DTP serologies and doing the pertussis antibody tests; Dr. Meade for editing, assisting with analysis and interpretation of serologic data; Dr. Collins for supervising the malaria antibody testing; Dr. Lobel for assisting with the protocol, field work and expedition of the serologic analyses; Dr. Saliou for participating in field work and manuscript analysis; Ms. Roberts for data registration and preliminary analysis; Dr. Campaoré for serving as the responsible health officer in Burkina Faso; and Dr. Miller for analysis and interpretation of results.
Financial Support
The study was supported partly by a grant from the Malaria Unit at the World Health Organization. Amodiaquine was contributed by Warner Lambert Pharmaceuticals.
Acknowledgements
We thank Kenneth Herrmann, Barbara Dove, John Robbins, James Nakano (deceased), Henry Mathews, Allen Hightower, Peggy Stanfil, Lois Norman (deceased), Cecile Viboud, Juan Arciniega, and Vicki Breman for their contributions and Cherice Holloway for manuscript assistance. We are deeply indebted to the children and the families who participated in the study.
Drs. Rosen and Breman are co-first authors. Dr. Breman was with the Centers for Disease Control and Prevention, and he and Dr. Saliou were assigned to the Centre Muraz, Organisation de Coordination et de Coopération pour la lutte contre les Grandes Endémies when this study was performed.
==== Refs
Breman JG Ailio MS Mills A Conquering the intolerable burden of malaria: what's new, what's needed: a summary American Journal of Tropical Medicine and Hygiene 2004 71(Suppl 2) 1 15 15331814
Bryce J Boschi-Pinto C Shibuya K Black RE WHO estimates of the causes of death in children Lancet 2005 365 1147 1152 15794969 10.1016/S0140-6736(05)71877-8
Geerligs PD Brabin BJ Eggelte TA Analysis of the effects of malaria chemoprophylaxis in children on haematological responses, morbidity and mortality Bull World Health Organ 2003 81 205 216 12764517
Massaga JJ Kitua AY Lemnge MM Akida JA Malle LN Ronn AM Theander TG Bygbjerg IC Effect of intermittent treatment with amodiaquine on anaemia and malarial fevers in infants in Tanzania: a randomised placebo-controlled trial Lancet 2003 361 1853 1860 12788572 10.1016/S0140-6736(03)13504-0
Schellenberg D Menendez C Kahigwa E Aponte J Vidal J Tanner M Mshinda H Alonso P Intermittent treatment for malaria and anaemia control at time of routine vaccinations in Tanzanian infants: a randomised, placebo-controlled trial Lancet 2001 357 1471 1477 11377597 10.1016/S0140-6736(00)04643-2
Rosen JR Breman JG Malaria Intermittent Preventive Treatment in Infants (IPTi), Chemoprophylaxis, and Childhood Vaccinations Lancet 2004 363 1386 1388 15110499 10.1016/S0140-6736(04)16052-2
Schellenberg D Menendez C Aponte JJ Kahigwa E Tanner M Mshinda H alonso P Intermittent preventive antimalarial treatment for Tanzanian infants: follow-up to age 2 years of a randomized placebo-controlled trial Lancet 2005 365 1481 1483 15850632 10.1016/S0140-6736(05)66418-5
Greenwood BM Bradley-Moore AM Bryceson AD Palit A Immunosuppression in children with malaria Lancet 1972 1 169 172 4109547 10.1016/S0140-6736(72)90569-7
Williamson WA Greenwood BM Impairment of the immune response to vaccination after acute malaria Lancet 1978 1 1328 1329 78096 10.1016/S0140-6736(78)92403-0
Usen S Milligan P Ethevenaux C Greenwood B Mulholland K Effect of fever on the serum antibody response of Gambian children to Haemophilus influenzae type b conjugate vaccine Pediatr Infect Dis J 2000 19 444 449 10819341 10.1097/00006454-200005000-00010
Tarzaali A Viens P Quevillon M Inhibition of the immune response to whooping cough and tetanus vaccines by malaria infection, and the effect of pertussis adjuvant Am J Trop Med Hyg 1977 26 520 524 869103
Greenwood BM Bradley AK Blakebrough IS Whittle HC Marshall TF Gilles HM The immune response to a meningococcal polysaccharide vaccine in an African village Trans R Soc Trop Med Hyg 1980 74 340 346 6776665 10.1016/0035-9203(80)90095-4
Simondon F Preziosi MP Pinchinat S Yam A Chabirand L Wassilak S Pines E Trape JF Salomon H Hoffenbach A Randomised study of the possible adjuvant effect of BCG vaccine on the immunogenicity of diphtheria-tetanus-acellular pertussis vaccine in Senegalese infants Eur J Clin Microbiol Infect Dis 1999 18 23 29 10192710 10.1007/s100960050221
Monjour L Bourdillon F Schlumberger M Fayet MT Michon C Ballet JJ Gouba E Gentilini M [Humoral and cellular immunity following antitetanus vaccination in malnourished and malaria-induced African children. 1. Study of the antitetanus antibody response] Bull World Health Organ 1982 60 589 596 6754123
Monjour L Palminteri R Froment A Renault T Alfred C Gentilini M Gouba E Is cell-mediated immune response related to nutritional state, but unaffected by concomitant malarial infection? Ann Trop Med Parasitol 1982 76 575 577 7159110
Cenac A Develoux M Djibo A Chloroquine treatment of malaria does not increase antibody response to measles vaccination. A controlled study of 580 rural children living in an endemic malaria area Trans R Soc Trop Med Hyg 1988 82 405 3068857 10.1016/0035-9203(88)90135-6
Bradley-Moore AM Greenwood BM Bradley AK Bartlett A Bidwell DE Voller A Craske J Kirkwood BR Gilles HM Malaria chemoprophylaxis with chloroquine in young Nigerian children. II. Effect on the immune response to vaccination Ann Trop Med Parasitol 1985 79 563 573 3834841
Greenwood AM Greenwood BM Bradley AK Enhancement of the immune response to meningococcal polysaccharide vaccine in a malaria endemic area by administration of chloroquine Ann Trop Med Parasitol 1981 75 261 263
Spindel R Baruzzi RG Souza VA Ferreira AW Avila SL Measles vaccine coverage and immune response in children of Caiabi and Metuktire Indian tribes living in malarial endemic area: Parque indigena do Xingu, Central Brazil Trop Doct 2001 31 142 144 11444333
McGregor I Barr M Antibody response to tetanus toxoid innoculation in malarious and non-malarious Gambian children Trans R Soc Trop Med Hyg 1962 56 364 367 10.1016/0035-9203(62)90005-6
Kollaritsch H Que JU Kunz C Wiedermann G Herzog C Cryz SJJ Safety and immunogenicity of live oral cholera and typhoid vaccines administered alone or in combination with antimalarial drugs, oral polio vaccine, or yellow fever vaccine J Infect Dis 1997 175 871 875 9086143
Fryauff DJ Church LW Richards AL Widjaja H Mouzin E Ratiwayanto S Hadiputranto H Sutamihardja MA Richie TL Subianto B Tjitra E Hoffman SL Lymphocyte response to tetanus toxoid among Indonesian men immunized with tetanus-diphtheria during extended chloroquine or primaquine prophylaxis J Infect Dis 1997 176 1644 1648 9395384
Taylor DN Wasi C Bernard K Chloroquine prophylaxis associated with a poor antibody response to human diploid cell rabies vaccine Lancet 1984 1 1405 6145850 10.1016/S0140-6736(84)91893-2
Pappaioanou M Fishbein DB Dreesen DW Schwartz IK Campbell GH Sumner JW Patchen LC Brown WJ Antibody response to preexposure human diploid-cell rabies vaccine given concurrently with chloroquine N Engl J Med 1986 314 280 284 3510393
Escodie A Hamon J Le paludisme en Afrique occidentale d'expression francaise Med Trop 1961 21 661 687
www.who.int/vaccines/globalsummary/immunization/countryprofileresult.cfm
Krishna S White NJ Pharmacokinetics of quinine,
chloroquine and amodiaquine. Clinical implications Clinical Pharmacokinetics 1996 30 263 269 8983859
Winstanley PA Simooya O Kofi-Ekue JM Oalker O Salako LA Edwards G Orme ML Breckenridge AM The disposition of amodiaquine in Zambians and Nigerians with malaria British Journal of Pharmacology 1990 29 695 701
Watkins WM Sixsmith DG Spencer HC Boriga DA Kariuki DM Kipingor T Koech DK Effectiveness of amodiaquine as treatment for chloroquine-resistant Plasmodium falciparum infections in Kenya Lancet 1984 1 357 359 6141423 10.1016/S0140-6736(84)90410-0
Mutabingwa TK Anthony D Heller A Hallett R Ahmed J Drakeley C Greenwood B Whitty CJM Amodiaquine alone, amodiaquine + sulfadoxine-pyrimethamine, amodiaquine + artesunate, and artemether-lumefantrine for outpatient treatment of malaria in Tanzanian children: a four-arm randomized effectiveness trial Lancet 2005 365 1474 1480 15850631 10.1016/S0140-6736(05)66417-3
Hierholzer JC Suggs MT Hall EC Standardized viral hemagglutination and hemagglutination-inhibition tests: II. Description and statistical evaluation Applied Microbiology 1969 18 824 833 4984203
Hierholzer JC Suggs MT Standardized viral hemagglutination and hemagglutination-inhibition tests: I. Standardization of erythrocyte suspension Applied Microbiology 1969 18 816 823 4984202
Wassilak SG Bernier RH Herrmann KL Orenstein WA Bart KJ Amler R Measles seroconfirmation using dried capillary blood specimens in filter paper Pediatric Infectious Disease 1984 3 117 121 6728701
Boyden SV The adsorption of proteins on erythrocytes treated with tannic acid and subsequent hemagglutination by antiprotein sera J Exp Med 1951 93 107 120 14803635 10.1084/jem.93.2.107
Stavitsky AB Micromethods for the study of protein and antibodies. I. Procedure and general applications of hemagglutination and hemagglutination-inhibition reactions with tannic acid and protein-treated red blood cells J Immunol 1954 72 360 367 13163409
Manclark CR Serological Response to Bordetella pertussis Manual of Clinical Immunology 1976 Baltimore, The Williams and Wilkins Co. 312 314
Sulzer AJ Wilson M Hall EC Indirect fluorescent antibody test for parasitic diseases. V. An evaluation of a thick-smear antigen in the IFA test for malaria antibodies Am J Trop Med Hyg 1969 18 199 205 4888028
Spencer HC Collins WE Chin W Skinner JC The Enzyme-linked immunosorbent assay (ELISA) for malaria I. The use of in vitro-cultured Plasmodium falciparum as antigen Am J Trop Med Hyg 1979 28 927 932 116555
Scorza T Magez S Brys L De Baetselier P Hemozoin is a key factor in the induction of malaria-associated immunosuppression Parasite Immunol 1999 21 545 554 10583855 10.1046/j.1365-3024.1999.00254.x
McBride JS Micklem HS Immunosuppression in murine malaria. II. The primary response to bovine serum albumin Immunology 1977 33 253 259 608688
Schwarzer E Bellomo G Giribaldi G Ulliers D Arese P Phagocytosis of malarial pigment haemozoin by human monocytes: a confocal microscopy study Parasitology 2001 123 125 131 11510677 10.1017/S0031182001008216
Ocana-Morgner C Mota MM Rodriguez A Malaria blood stage suppression of liver stage immunity by dendritic cells J Exp Med 2003 197 143 151 12538654 10.1084/jem.20021072
Ho M Webster HK Looareesuwan S Supanaranond W Phillips RE Chanthavanich P Warrell DA Antigen-specific immunosuppression in human malaria due to Plasmodium falciparum J Infect Dis 1986 153 763 771 2936834
Urban BC Ferguson DJ Pain A Willcox N Plebanski M Austyn JM Roberts DJ Plasmodium falciparum-infected erythrocytes modulate the maturation of dendritic cells Nature 1999 400 73 77 10403251 10.1038/21900
Urban BC Willcox N Roberts DJ A role for CD36 in the regulation of dendritic cell function Proc Natl Acad Sci U S A 2001 98 8750 8755 11447263 10.1073/pnas.151028698
Rockett KA Awburn MM Rockett EJ Cowden WB Clark IA Possible role of nitric oxide in malarial immunosuppression Parasite Immunol 1994 16 243 249 8072768
Snyder DS Beller DI Unanue ER Prostaglandins modulate macrophage Ia expression Nature 1982 299 163 165 6287286 10.1038/299163a0
Mabey DC Brown A Greenwood BM Plasmodium falciparum malaria and Salmonella infections in Gambian children J Infect Dis 1987 155 1319 1321 3553352
Gunapala DE Facer CA Davidson R Weir WR In vitro analysis of Epstein-Barr virus: host balance in patients with acute Plasmodium falciparum malaria. I. Defective T-cell control Parasitol Res 1990 76 531 535 2166284 10.1007/BF00931060
Cook IF Herpes zoster in children following malaria J Trop Med Hyg 1985 88 261 264 3910848
Clark IA al-Yaman FM Cowden WB Rockett KA Does malarial tolerance, through nitric oxide, explain the low incidence of autoimmune disease in tropical Africa? Lancet 1996 348 1492 1494 8942781 10.1016/S0140-6736(96)07342-4
Adebajo AO Low frequency of autoimmune disease in tropical Africa Lancet 1997 349 361 362 9024407 10.1016/S0140-6736(05)62867-X
Wangoo A Ganguly NK Mahajan RC Immunosuppression in murine malaria: suppressor role of macrophages and their products during acute and chronic Plasmodium berghei infection Apmis 1990 98 407 414 2141476
Plebanski M Flanagan KL Lee EA Reece WH Hart K Gelder C Gillespie G Pinder M Hill AV Interleukin 10-mediated immunosuppression by a variant CD4 T cell epitope of Plasmodium falciparum Immunity 1999 10 651 660 10403640 10.1016/S1074-7613(00)80064-3
Feeney C Bryzman S Kong L Brazil H Deutsch R Fritz LC T-lymphocyte subsets in acute illness Crit Care Med 1995 23 1680 1685 7587233 10.1097/00003246-199510000-00012
Shimizu Y Newman W Gopal TV Horgan KJ Graber N Beall LD van Seventer GA Shaw S Four molecular pathways of T cell adhesion to endothelial cells: roles of LFA-1, VCAM-1, and ELAM-1 and changes in pathway hierarchy under different activation conditions J Cell Biol 1991 113 1203 1212 1710227 10.1083/jcb.113.5.1203
Tanaka Y Adams DH Hubscher S Hirano H Siebenlist U Shaw S T-cell adhesion induced by proteoglycan-immobilized cytokine MIP-1 beta Nature 1993 361 79 82 7678446 10.1038/361079a0
Monjour L Bourdillon F Froment A Daniel-Ribeiro C Tirard S Datry A Chastang C Tselentis Y Gentilini M [Measles vaccination in the Sudan-Sahel region of Africa. Absence of the immunodepressive effect of malaria] Pathol Biol (Paris) 1985 33 232 235 3892451
Meade BD Lynn F Reed GF Mink CM Romani TA Deforest A Deloria MA Relationships between functional assays and enzyme immunoassays as measurements of responses to acellular and shole-cell pertussis vaccines Pediatrics 1995 96 595 600 7659484
Vaccination against whooping cough: relation between protection in children and results for laboratory tests; a report to the Whooping-cough Immunization Committee of the Medical Research Council and to the medical officers of health for Cardiff, Leeds, Leyton, Manchester, Middlesex, Oxford, Poole, Tottenham, Walthamstow, and Wembley British Medical Journal 1956 2 454 651 13342513
Durbaca S Antidiphtheria and antitetanus immunity of recruits in Romania Roumanian archives of microbiology and immunology 1996 55 295 303 9558964
|
16271153
|
PMC1308854
|
CC BY
|
2021-01-04 16:37:31
|
no
|
Malar J. 2005 Nov 6; 4:53
|
utf-8
|
Malar J
| 2,005 |
10.1186/1475-2875-4-53
|
oa_comm
|
==== Front
Malar JMalaria Journal1475-2875BioMed Central London 1475-2875-4-541627765410.1186/1475-2875-4-54ResearchComplement receptor 1 polymorphisms associated with resistance to severe malaria in Kenya Thathy Vandana [email protected] JoAnn M [email protected] Bernard [email protected] Walter [email protected] José A [email protected] The US Army Medical Research Unit and the Kenya Medical Research Institute, Nairobi, Kenya2 Department of Microbiology and Immunology, Drexel University College of Medicine, Philadelphia, PA, USA3 Department of Cellular Injury, Walter Reed Army Institute of Research, Silver Spring, MD and Department of Medicine, Uniformed Services University of Health Sciences, Bethesda, MD, USA4 LifeShare Blood Centers, Shreveport, LA, USA2005 8 11 2005 4 54 54 9 8 2005 8 11 2005 Copyright © 2005 Thathy et al; licensee BioMed Central Ltd.2005Thathy et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
It has been hypothesized that the African alleles Sl2 and McCb of the Swain-Langley (Sl) and McCoy (McC) blood group antigens of the complement receptor 1 (CR1) may confer a survival advantage in the setting of Plasmodium falciparum malaria, but this has not been demonstrated.
Methods
To test this hypothesis, children in western Kenya with severe malaria-associated anaemia or cerebral malaria were matched to symptomatic uncomplicated malaria controls by age and gender. Swain-Langley and McCoy blood group alleles were determined by restriction fragment length polymorphism and conditional logistic regression was carried out.
Results
No significant association was found between the African alleles and severe malaria-associated anaemia. However, children with Sl2/2 genotype were less likely to have cerebral malaria (OR = 0.17, 95% CI 0.04 to 0.72, P = 0.02) than children with Sl1/1. In particular, individuals with Sl2/2 McCa/b genotype were less likely to have cerebral malaria (OR = 0.18, 95% CI 0.04 to 0.77, P = 0.02) than individuals with Sl1/1 McCa/a.
Conclusion
These results support the hypothesis that the Sl2 allele and, possibly, the McCb allele evolved in the context of malaria transmission and that in certain combinations probably confer a survival advantage on these populations.
==== Body
Background
Plasmodium falciparum malaria is responsible for most of the more than one million deaths that occur each year from malaria infection in Africa [1]. Most of these deaths occur as a result of complications such as severe malaria-associated anaemia (SMA) and cerebral malaria (CM) or coma [2]. The pathogenesis of these complications is poorly understood. Evidence from several studies [3-7] suggests that the complement receptor 1 (CR1, CD35) may be involved in the pathogenesis of severe malaria. However, its exact role is not known.
CR1 is a protein ranging in Mr from 190 to 280 kDa. It is found on erythrocytes and most leukocytes and it is divided into three to four long homologous repeat regions (LHRs) and 27 to 30 short consensus repeats (SCRs), also known as complement control protein repeats (CCPs) [8] (Figure 1). LHR-A binds C4b, whereas LHR-B and C bind both C4b and C3b. LHR-D binds mannan-binding lectin and C1q [9,10]. Together with other erythrocyte complement regulatory proteins such as decay accelerating factor (DAF, CD55) and membrane inhibitor of reactive lysis (MIRL, CD59), CR1 serves to attenuate the complement cascade. Specifically, CR1 promotes the Factor I-mediated inactivation of C3b to C3bi and promotes inactivation of C3 convertases. CR1 on erythrocytes also binds opsonized immune complexes carrying them to organs of the reticuloendothelial system for removal. Altogether, these functions serve to protect erythrocytes and other host cells from complement-mediated damage. In addition, CR1 is thought to be a receptor for the P. falciparum erythrocyte membrane protein 1 (PfEMP-1) [3,4]. The interaction between PfEMP-1 and CR1 may be responsible for rosetting, a phenomenon in which erythrocytes infected with P. falciparum late stage parasites bind to uninfected erythrocytes in vitro and has been associated with cerebral malaria in several African studies [11,12].
Figure 1 Schematic diagram of the most common structural variant of complement receptor 1 (CR1*1, F allele). The amino terminal (NH2) extracellular portion is composed of 30 complement control protein repeats (CCPs; vertical boxes numbered 1–30) arranged into four long homologous regions (LHRs) A-D, each composed of seven CCPs. There are two distinct functional domains each composed of three complement control protein repeats (CCPs) (vertical hatched boxes): site 1 in LHR-A (CCPs 1–3) binds mainly C4b and has convertase decay accelerating activity, and two virtually identical copies of site 2 in LHR-B (CCPs 8–10) and LHR-C (CCPs 15–17) that bind C3b and C4b, as well as PfEMP-1, and possess Factor I- cofactor activity. Functional differences in sites 1 and 2 are determined by amino acid sequence differences. Boxes with the same hatching pattern reflect near amino acid identity of the CCPs. CCP 25 (stippled box) carries the Swain-Langley (Sl) and McCoy (McC) Knops blood group antigens. The CCP repeats are followed by a transmembrane domain (TM) and a cytoplasmic tail (CYT).
The CR1 gene exhibits a number of polymorphisms. There are several size variants which are felt to be the result of unequal gene crossover [13]. The most common variant has four LHRs (Figure 1). Polymorphisms that correlate with the quantitative expression of CR1 on erythrocytes of Caucasians but not of Africans have also been identified [14,15]. Lastly, the Knops blood group antigens, including the Swain-Langley (Sl), McCoy (McC) and their alleles, have been localized to the CR1 protein in LHR-D [16-18].
The study of the Sl and McC blood group antigens may offer further clues as to the role of CR1 in the pathogenesis of severe malaria. The phenotypic absence of these antigens, designated Sl -1,2 [19] and McC(a-b+) respectively, is occasionally the result of low CR1 copy numbers on the RBC or, most commonly, due to non-synonymous base substitutions [17]. The alleles responsible for the phenotypic absence are very rare in ethnic groups that are not of African descent. Thus, this observation has led to the hypothesis that the Sl2 and McCb alleles may confer a survival advantage in the face of diseases such as malaria. In support of this hypothesis Sl -1,2 erythrocytes were found to rosette less in vitro than Sl 1,-2 erythrocytes [3]. However, a recent study failed to reveal an association between Sl2 or McCb and resistance to severe malaria in The Gambia, West Africa [20]. The present study describes the findings of an investigation designed to determine whether there is an association between resistance or susceptibility to severe malaria and the Sl and McC blood group alleles in western Kenya, a region of malaria endemicity. The findings suggest that, contrary to a previous report [20], Sl2 does confer an advantage in the setting of malaria infection.
Methods
Populations and study design
Participants were recruited under human use protocols approved by the Human Subjects Research Review Board, Office of the Surgeon General, US Army, and the Ethics Review Committee of the Kenya Medical Research Institute, Nairobi, Kenya. All procedures were in accordance with the Helsinki Declaration. The study had a matched case-control design. The demographics of some of the study participants as well as the inclusion/exclusion criteria have been described before [5-7]. SMA cases, defined as children with asexual P. falciparum parasitaemia by Giemsa-stained thick and thin blood smear and Hb ≤ 5 g/dL, were recruited from the pediatric ward of the Nyanza Provincial General Hospital (NPGH), Kisumu, Kenya. The NPGH catchment area is the malaria holoendemic region of the Lake Victoria basin in western Kenya, where most individuals are of the Luo ethnic group. Because CM is uncommon in the Lake Victoria basin, CM cases were recruited from the pediatric ward of the Kisii District Hospital (KDH), as well as from the NPGH. KDH is located in the highlands of western Kenya where transmission is seasonal and consequently receives many more CM cases than the NPGH [21]. The predominant ethnic group in Kisii is the Abagusii. CM was defined as asexual P. falciparum parasitaemia by Giemsa-stained blood smear and a Blantyre coma score of ≤ 2 [22], lasting at least 30 min if there was a history of convulsions. One control with symptomatic uncomplicated malaria was matched to each case by age ± two months and gender and was identified from the outpatient clinic of the same hospital where the corresponding case was obtained. Symptomatic uncomplicated malaria was defined as a Giemsa-stained blood smear positive for asexual P. falciparum and an axillary temperature ≥ 37.5°C or, in the absence of the latter, two of the following signs or symptoms: nausea/vomiting, irritability, poor feeding, myalgias or headache. Children were excluded from participation if there was evidence of other concomitant serious infections (i.e. meningitis excluded by lumbar puncture, pneumonia, sepsis) or chronic illness. All children with suspected CM underwent a lumbar puncture to exclude meningitis and were enrolled if the results of Gram stain and culture were negative. In addition, because some of our studies also included the measurement of erythrocyte complement regulatory proteins, children were excluded if they had a history of blood transfusion in the three months preceding enrolment to avoid the influence of donor erythrocytes.
Genotyping of CR1 Knops blood group polymorphisms
A whole blood sample obtained at enrolment, and prior to any blood transfusion, was used to extract DNA and/or blotted onto filter papers for later extraction. Personnel who were unaware of the group assignments of the study participants carried out DNA extraction, amplification and interpretation of restriction fragment length polymorphisms (RFLPs) patterns. DNA was extracted from whole blood using the QIAamp DNA blood extraction protocol (Qiagen, Valencia, CA). DNA extraction from filter papers was performed using saponin and Chelex as previously described [23]. Once extracted, genomic DNA was stored frozen at -20°C for later analysis. The Sl1/Sl2 and McCa/McCb Knops blood group alleles result from single nucleotide polymorphisms that cause specific amino acid changes[17] in exon 29 (encoding CCP 25; Figure 1). An established PCR-RFLP method was used for the detection of both SNPs [24]. Briefly, the oligonucleotide forward primer 24KnNde: 5'-ACC AGT GCC ACA CTG GAC CAG ATG GAG AAC AGC TGT TTG AGC AT-3' and reverse primer 25Rb: 5'-GGA GGA GTG TGG CAG CTT G-3', were used to amplify a 305 bp fragment of CR1 exon 29 by PCR. The amplification reactions were carried out on ~500 ng of genomic DNA in a 100 μl final volume containing 1× GeneAmp PCR Buffer II (Applied Biosystems, Foster City, CA), 2.5 mM MgCl2, 200 μM each dNTP, 0.2 μM each primer and 5 U AmpliTaq Gold® DNA Polymerase (Applied Biosystems, Foster City, CA). After 9 min of initial denaturation at 95°C, the thermocycling program used consisted of 1 min denaturation at 94°C, 1 min annealing at 58°C, and 1 min extension at 72°C for 44 cycles, followed by a final extension of 10 min at 72°C. For RFLP analysis, 28 μL of each PCR product was digested with either MfeI (New England Biolabs, Beverly, MA), for the detection of Sl1/Sl2, or BsmI (New England Biolabs), for the detection of McCa/McCb, and the restriction fragments were resolved by agarose gel electrophoresis as previously described [24].
Statistical analysis
All analyses were done using SPSS version 11.5 (SPSS Inc., Chicago, IL). Chi-square tests were performed to compare the observed and predicted frequencies of genotypes based on allelic frequencies in different ethnic groups. Conditional logistic regression based on matched pairs and adjusted for ethnic groups was carried out for each single locus (Sl and McC) and for the combination of the two loci. Thus, each regression model included terms both for genotype and ethnic group. All tests were two-sided with α = 0.05.
Results
Population demographics
Table 1 describes the total numbers of cases and controls per group and hospital. The cases and controls were well matched in age, gender and ethnic group distribution. Most of the cases and controls recruited from the NPGH were of the Luo ethnic group, whereas most of the cases and controls recruited from the KDH were of the Abagusii ethnic group.
Table 1 Group demographics
Nyanza Provincial General Hospital N = 320 Kisii District Hospital N = 140
Severe Anaemia Cerebral Malaria Cerebral Malaria
Cases N = 137 Controls N = 137 Cases N = 23 Controls N = 23 Cases N = 70 Controls N = 70
Mean age in months (SD) 14.8(12.6) 14.5(12.2) 29.0(15.7) 30.1(15.7) 28.5(15.9) 29.0(15.8)
No. of females (%) 57(41.6) 57(41.6) 9(39.1) 9(39.1) 37(52.9) 37(52.9)
Ethnic groups (%)
Luo 114(83.2) 115(83.9) 22(95.7) 19(82.6) 3(4.3) 4(5.7)
Abagusii 1(0.7) 0 1(4.3) 4 (17.4) 66(94.3) 65(92.9)
Luhya 15(10.9) 19(13.9) 0 0 1(1.4) 1(1.4)
Other 7(5.1) 3(2.2) 0 0 0 0
Other: Kalenjin (3), Kikuyu (3), Turkana (1), Teso (1), Nandi (1), Bukusu (1)
Genotype frequencies
Sl2 and McCb alleles were more common in The Gambia [20], Mali [17] and western Kenya than in other non-African ethnic or racial groups where they occurred in very few numbers (Table 2). There were no significant differences between predicted and observed genotype frequencies for the two major ethnic groups, Luos and Abagusiis (data not shown), suggesting that the genotypes within these populations are in Hardy-Weinberg equilibrium.
Table 2 Percent of Swain-Langley and McCoy genotypes in various populations.
Populations Swain-Langley McCoy Reference
1/1 1/2 2/2 a/a a/b b/b
Western Kenya N = 460 10 44 45 48 45 7 This paper
The Gambia N = 853 5 31 65 38 47 15 [20]
Mali N = 99 9 30 60 49 40 10 [17]
Caucasian Americans N = 100 99 1 0 100 0 0 [20]
Asian Americans N = 99 95 4 1 96 4 0 [20]
Hispanic Americans N = 100 94 6 0 95 5 0 [20]
Sl2/2 is associated with decreased susceptibility to cerebral malaria
In order to explore whether the Sl2 or McCb alleles confer resistance or susceptibility to severe malaria compared to the Sl1 and McCa alleles, conditional logistic regression of the individual genotypes at each locus was carried out using Sl1/1 or McCa/a as reference (Table 3). There was no association between the Sl1/2 or Sl2/2 genotype and resistance or susceptibility to SMA. Although there was a trend towards decreased susceptibility to SMA for the Sl2/2 genotype, the association was not significant (OR = 0.65, 95% CI 0.14 to 3.06, P = 0.59). On the other hand, individuals with the Sl2/2 genotype were less likely to have cerebral malaria than individuals with Sl1/1 (OR = 0.17, 95% CI 0.04 to 0.72, P = 0.02). This finding was independent of the hospital site since inclusion of this variable in the model did not affect the results. There was no significant effect of the McCa/b or McCb/b genotypes, when averaged over the different Sl alleles, on resistance or susceptibility to CM.
Table 3 Conditional logistic regression based on matching to compare individual genotypes.
Severe Anaemia Cerebral Malaria
Genotype No. of cases (%) No. of controls (%) OR 95% CI P No. of cases (%) No. of controls (%) OR 95% CI P
Sl1/1 11(8) 10(7) Ref - - 17(18) 10(11) Ref - -
Sl1/2 63(46) 57(42) 1.08 0.24 to 4.93 0.92 44(47) 40(43) 0.49 0.12 to 2.05 0.32
Sl2/2 63(46) 70(51) 0.65 0.14 to 3.06 0.59 32(34) 43(46) 0.17 0.04 to 0.72 0.02
McCa/a 61(45) 61(45) Ref - - 51(55) 49(53) Ref - -
McCa/b 65(47) 66(48) 0.93 0.41 to 2.11 0.86 37(40) 40(43) 1.39 0.55 to 3.52 0.49
McCb/b 11(8) 10(7) 1.38 0.29 to 6.57 0.68 5(5) 4(4) 3.52 0.43 to 29.06 0.24
OR = Odds Ratio, CI = Confidence interval, Ref = Reference Category
Sl2/2 McCa/b genotype is associated with decreased susceptibility to cerebral malaria
In order to determine the effect of the different genotype combinations at the Swain-Langley and McCoy loci on the susceptibility to severe malaria conditional logistic regression was carried out using the Sl1/1 McCa/a genotype as reference (Table 4). Only six of the ten possible Sl/McC genotype combinations were found. Although it is not known whether the haplotypes for the group with genotype Sl1/2 McCa/b are 1a/2b or 1b/2a, it was assumed that these individuals were 1a/2b because genotypes Sl1/1 McCa/b, Sl1/1 McCb/b and Sl1/2 McCb/b were not found in the population, suggesting that haplotype 1b is extremely rare or does not exist. Table 4 shows that among CM cases and controls with McCa/a background the susceptibility to CM (OR) decreased from Sl1/2 to Sl2/2, suggesting a gene dose effect. The strongest association was seen with genotype Sl2/2 McCa/b (OR = 0.18, 95% CI = 0.04 to 0.77, P = 0.02). Interestingly, heterozygosity at both loci or homozygozity for Sl2 and McCb did not show any significant benefit.
Table 4 Conditional logistic regression to compare genotype combinations to the non-African genotype Sl1/1 McCa/a.
Severe Anaemia Cerebral Malaria
Genotype No. of cases (%) No. of controls (%) OR 95% CI P No. of cases (%) No. of controls (%) OR 95% CI P
Sl1/1 McCa/a 11(8) 10(7) Ref - - 17(18) 10(11) Ref - -
Sl1/2 McCa/a 33(24) 34(25) 0.93 0.20 to 4.37 0.93 23(25) 25(27) 0.40 0.09 to 1.77 0.23
Sl2/2 McCa/a 17(12) 17(12) 0.86 0.17 to 4.43 0.86 11(12) 14(15) 0.24 0.05 to 1.19 0.08
Sl2/2 McCa/b 35(26) 43(31) 0.53 0.12 to 2.34 0.40 16(17) 25(27) 0.18 0.04 to 0.77 0.02
Sl1/2 McCa/b 30(22) 23(17) 1.19 0.26 to 5.40 0.82 21(23) 15(16) 0.93 0.19 to 4.60 0.93
Sl2/2 McCb/b 11(8) 10(7) 0.87 0.15 to 5.14 0.88 5(5) 4(4) 0.65 0.07 to 6.07 0.71
OR = Odds Ratio, CI = Confidence Interval, Ref = Reference category
Discussion
The present study has shown that individuals with the Sl2/2 genotype that confers the Sl -1,2 phenotype have decreased susceptibility to CM compared to individuals with the Sl1/1 genotype. This finding was independent of ethnicity, of the genotype at the McC locus and of the hospital site where the cases and controls were recruited. Among the combined genotypes, a trend of increasing benefit was observed in the background of McCa/a from Sl1/2 to Sl2/2, which is consistent with a gene dose effect. Although the level of statistical significance may be influenced by a larger sample size in the Sl2/2 category, the association nevertheless appears strongest with the acquisition of one McCb allele (Table 4).
The finding that genotype Sl2/2 McCa/b might be the one that most influences the level of susceptibility to CM may seem counterintuitive since in Table 3 McCa/b did not show any significant advantage and perhaps showed a trend towards a detrimental effect relative to McCa/a. However, in the analysis in Table 3 the reference genotype McCa/a contains three variants: Sl1/1 McCa/a, Sl1/2 McCa/a, and Sl2/2 McCa/a with the latter two both showing indications of decreased susceptibility due to the presence of Sl2 (Table 4). The lack of significant benefit from Sl2/2 in SMA probably reflects the distinct pathogenic mechanisms between this condition and CM, but this does not exclude a role for CR1 in the pathogenesis of SMA.
In contrast to this study, a previous case-control study in The Gambia [20] failed to demonstrate any significant association between Sl2/2 and resistance to severe malaria. The divergence in findings between the present and the previous study could be explained by important methodological differences. In the Gambian study controls did not have malaria. This may have led to an overrepresentation of non-protective genotypes in the controls because these children did not have the most important risk factor for severe malaria, i.e. malaria. Further, cases and controls were not matched by age in the Gambian study, and in fact, cases were older than controls by an average of almost one year. In addition, it is possible that differences in the pathogenesis of severe malaria between the two sites on the basis of other genetic factors and/or differences in the malaria transmission patterns may exist.
The mechanism by which the Sl2 and McCb alleles could confer protection from severe malaria is not known, since these polymorphisms are located near the C-terminus of the CR1 molecule in an area that is outside the binding sites for C3b, C4b, and PfEMP-1. Nonetheless, there are a number of potential explanations based on the knowledge of CR1 function and the nature of the amino acid substitutions that take place. The Sl2 allele is the result of the substitution of glycine, a neutral amino acid, for the basic amino acid arginine at position 1601 (R1601G), whereas the McCb allele is the result of the substitution of the acidic amino acid glutamic acid for the basic amino acid lysine at position 1590 (K1590E) [17]. These modifications may inflict significant conformational or charge changes to the rest of the molecule and may consequently impact on the conformation and function of the C3b and C4b binding sites and/or the PfEMP-1 binding site. Accordingly, Sl -1,2 erythrocytes have been reported to rosette less than Sl 1,-2 erythrocytes [3]. In addition, it is known that CR1 molecules are aggregated on the erythrocyte surface and this aggregation is felt to be critical to the binding affinity of CR1 for C3b and C4b [25]. Therefore, it is possible that the amino acid substitutions not only induce conformational changes to the individual CR1 molecules but also affect the ability of functional aggregates to form due to conformational incompatibility or charge interference between different CR1 molecules. This could explain why some combinations of these alleles, such as heterozygosity at both loci or homozygosity for both Sl2 and McCb, did not offer any significant advantage.
Although the current expectation is that CR1 may influence the development of severe malaria by virtue of its function as a regulator of the complement cascade on RBCs or by its direct interaction with the infected RBCs leading to rosette formation, CR1 may exert its effect on malaria in other ways that are not yet understood. In addition to being present on RBCs, CR1 is also present on follicular dendritic cells, macrophages and T and B lymphocytes and, therefore, may also have immunoregulatory functions that affect the development of immunity against malaria [13]. Mice deficient in CR1/CR2 have depressed humoral immune responses [26]. Further studies will be required to explore this possibility.
Further evidence that the role of CR1 in the pathogenesis of severe malaria may not be as expected is found in a recent case-control study in Papua New Guinea that aimed to determine whether there was an association between individuals who had the low CR1 expression allele (L) and resistance to severe malaria [27]. Contrary to expectation, the association was found only in heterozygotes (HL), and not in homozygotes (LL), which presumably had the lowest CR1 levels. In addition, although there was an association with resistance to severe malaria as a whole, there was no apparent association between these polymorphisms and CM. These results suggest that the relationship between the number of CR1 molecules on the RBC surface and severe malaria is not a straightforward one.
The results obtained in the study presented here have biological significance and are further supportive evidence of the importance of CR1 in the pathogenesis of severe malaria. Whatever the mechanism is by which CR1 is involved in the pathogenesis of malaria, knowledge of which alleles are associated with resistance or susceptibility to severe malaria will make it possible to observe the relationship between structure and function and susceptibility to severe malaria.
Conclusion
The results support the conclusion that the Sl2 allele and, possibly, the McCb allele have evolved in the context of malaria transmission in Africa and that in certain combinations probably confer a survival advantage to malaria endemic populations.
Authors' contributions
Vandana Thathy, assisted by Bernard Guyah, optimized and carried out the PCR, performed the restriction enzyme digestions and interpreted the RFLP results. JoAnn M. Moulds developed the PCR-RFLP assay. Walter Otieno supervised the clinical evaluations of the study participants. José A. Stoute was the principal investigator who designed the study, wrote the protocol and drafted the original manuscript that was reviewed and edited by all authors.
Acknowledgements
We would like to thank the children who participated in our studies as well as their parents for their willingness to contribute to this research endeavor. We are indebted to our dedicated staff of clinicians, nurses, drivers and field workers who made these studies possible. We are also grateful to Dr. John Rowlands for statistical support and for reviewing the paper. This work is published with the permission of the Office of the Director, the Kenya Medical Research Institute.
This work was supported in part by the US Army Medical Research and Materiel Command, by grants from the National Heart Lung and Blood Institute (RO1 HL 7502-01) to José A. Stoute and from the National Institute of Allergy and Infectious Diseases (RO1 AI 42367) to Joann M. Moulds. Vandana Thathy was supported by a fellowship from the Ellison Medical Foundation and the National Research Council. Bernard Guyah and Walter Otieno were supported by a training grant from the Fogarty International Center (1 D43 TW06239-01). Funding agencies had no role in the design of the study, the data collection, data analysis, data interpretation, writing of the manuscript, nor in the decision to submit the manuscript for publication.
"The views of the authors do not purport to reflect the position of the Department of the Army or the Department of Defense. The U.S. Government has the right to retain a nonexclusive, royalty-free license in and to any copyright covering this paper."
==== Refs
Organization WH World Health Report 2002 Geneva, WHO
Severe falciparum malaria. World Health Organization, Communicable Diseases Cluster Trans R Soc Trop Med Hyg 2000 94 Suppl 1 S1 90 11103309
Rowe JA Moulds JM Newbold CI Miller LH P. falciparum rosetting mediated by a parasite-variant erythrocyte membrane protein and complement-receptor 1 Nature 1997 388 292 295 9230440 10.1038/40269
Rowe JA Rogerson SJ Raza A Moulds JM Kazatchkine MD Marsh K Newbold CI Atkinson JP Miller LH Mapping of the region of complement receptor (CR) 1 required for Plasmodium falciparum rosetting and demonstration of the importance of CR1 in rosetting in field isolates J Immunol 2000 165 6341 6346 11086071
Waitumbi JN Opollo MO Muga RO Misore AO Stoute JA Red cell surface changes and erythrophagocytosis in children with severe Plasmodium falciparum anemia Blood 2000 95 1481 1486 10666228
Stoute JA Odindo AO Owuor BO Mibei EK Opollo MO Waitumbi JN Loss of red blood cell complement regulatory proteins and increased levels of circulating immune complexes are associated with severe malarial anemia J Infect Dis 2003 187 522 525 12552440 10.1086/367712
Waitumbi JN Donvito B Kisserli A Cohen JH Stoute JA Age-related changes in red blood cell complement regulatory proteins and the susceptibility to severe malaria J Infect Dis 2004 190 1183 1191 15319870 10.1086/423140
Krych-Goldberg M Atkinson JP Structure-function relationships of complement receptor type 1 Immunol Rev 2001 180 112 122 11414353 10.1034/j.1600-065X.2001.1800110.x
Ghiran I Barbashov SF Klickstein LB Tas SW Jensenius JC Nicholson-Weller A Complement receptor 1/CD35 is a receptor for mannan-binding lectin J Exp Med 2000 192 1797 1808 11120776 10.1084/jem.192.12.1797
Klickstein LB Barbashov SF Liu T Jack RM Nicholson-Weller A Complement receptor type 1 (CR1, CD35) is a receptor for C1q Immunity 1997 7 345 355 9324355 10.1016/S1074-7613(00)80356-8
Carlson J Helmby H Hill AV Brewster D Greenwood BM Wahlgren M Human cerebral malaria: association with erythrocyte rosetting and lack of anti-rosetting antibodies Lancet 1990 336 1457 1460 1979090 10.1016/0140-6736(90)93174-N
Rowe A Obeiro J Newbold CI Marsh K Plasmodium falciparum rosetting is associated with malaria severity in Kenya Infect Immun 1995 63 2323 2326 7768616
Ahearn JM Fearon DT Structure and function of the complement receptors, CR1 (CD35) and CR2 (CD21). Adv Immunol 1989 46 183 219 2551147
Herrera AH Xiang L Martin SG Lewis J Wilson JG Analysis of complement receptor type 1 (CR1) expression on erythrocytes and of CR1 allelic markers in Caucasian and African American populations Clin Immunol Immunopathol 1998 87 176 183 9614933 10.1006/clin.1998.4529
Rowe JA Raza A Diallo DA Baby M Poudiougo B Coulibaly D Cockburn IA Middleton J Lyke KE Plowe CV Doumbo OK Moulds JM Erythrocyte CR1 expression level does not correlate with a HindIII restriction fragment length polymorphism in Africans; implications for studies on malaria susceptibility Genes Immun 2002 3 497 500 12486610 10.1038/sj.gene.6363899
Moulds JM Nickells MW Moulds JJ Brown MC Atkinson JP The C3b/C4b receptor is recognized by the Knops, McCoy, Swain-langley, and York blood group antisera J Exp Med 1991 173 1159 1163 1708809 10.1084/jem.173.5.1159
Moulds JM Zimmerman PA Doumbo OK Kassambara L Sagara I Diallo DA Atkinson JP Krych-Goldberg M Hauhart RE Hourcade DE McNamara DT Birmingham DJ Rowe JA Moulds JJ Miller LH Molecular identification of Knops blood group polymorphisms found in long homologous region D of complement receptor 1 Blood 2001 97 2879 2885 11313284 10.1182/blood.V97.9.2879
Rao N Ferguson DJ Lee SF Telen MJ Identification of human erythrocyte blood group antigens on the C3b/C4b receptor J Immunol 1991 146 3502 3507 1827486
Daniels GL Cartron JP Fletcher A Garratty G Henry S Jorgensen J Judd WJ Levene C Lin M Lomas-Francis C Moulds JJ Moulds JM Moulds M Overbeeke M Reid ME Rouger P Scott M Sistonen P Smart E Tani Y Wendel S Zelinski T International Society of Blood Transfusion Committee on terminology for red cell surface antigens: Vancouver Report Vox Sang 2003 84 244 247 12670376 10.1046/j.1423-0410.2003.00282.x
Zimmerman PA Fitness J Moulds JM McNamara DT Kasehagen LJ Rowe JA Hill AV CR1 Knops blood group alleles are not associated with severe malaria in the Gambia Genes Immun 2003 4 368 373 12847553 10.1038/sj.gene.6363980
Marsh K Snow RW Malaria transmission and morbidity Parassitologia 1999 41 241 246 10697862
Molyneux ME Taylor TE Wirima JJ Borgstein A Clinical features and prognostic indicators in paediatric cerebral malaria: a study of 131 comatose Malawian children Q J Med 1989 71 441 459 2690177
Farnert A Arez AP Correia AT Bjorkman A Snounou G do Rosario V Sampling and storage of blood and the detection of malaria parasites by polymerase chain reaction Trans R Soc Trop Med Hyg 1999 93 50 53 10492790 10.1016/S0035-9203(99)90177-3
Moulds JM Thomas BJ Doumbo O Diallo DA Lyke KE Plowe CV Rowe JA Birmingham DJ Identification of the Kna/Knb polymorphism and a method for Knops genotyping Transfusion 2004 44 164 169 14962306 10.1111/j.1537-2995.2004.00615.x
Paccaud JP Carpentier JL Schifferli JA Difference in the clustering of complement receptor type 1 (CR1) on polymorphonuclear leukocytes and erythrocytes: effect on immune adherence Eur J Immunol 1990 20 283 289 2138082
Molina H Holers VM Li B Fung Y Mariathasan S Goellner J Strauss-Schoenberger J Karr RW Chaplin DD Markedly impaired humoral immune response in mice deficient in complement receptors 1 and 2 Proc Natl Acad Sci U S A 1996 93 3357 3361 8622941 10.1073/pnas.93.8.3357
Cockburn IA Mackinnon MJ O'Donnell A Allen SJ Moulds JM Baisor M Bockarie M Reeder JC Rowe JA A human complement receptor 1 polymorphism that reduces Plasmodium falciparum rosetting confers protection against severe malaria Proc Natl Acad Sci U S A 2004 101 272 277 14694201 10.1073/pnas.0305306101
|
16277654
|
PMC1308855
|
CC BY
|
2021-01-04 16:37:31
|
no
|
Malar J. 2005 Nov 8; 4:54
|
utf-8
|
Malar J
| 2,005 |
10.1186/1475-2875-4-54
|
oa_comm
|
==== Front
Mol CancerMolecular Cancer1476-4598BioMed Central London 1476-4598-4-391627447210.1186/1476-4598-4-39ResearchMapping and characterization of the amplicon near APOA2 in 1q23 in human sarcomas by FISH and array CGH Kresse Stine H [email protected] Jeanne-Marie [email protected] Leonardo A [email protected] Simon G [email protected] Wen-Lin [email protected] Joe W [email protected] Anne [email protected] Ola [email protected] Department of Tumour Biology, The Norwegian Radium Hospital, Oslo, Norway2 Faculty of Medicine, University of Oslo, Norway3 Center for Human Genetics, Duke University Medical Center, Durham, USA4 Comprehensive Cancer Centre, University of California San Francisco, USA5 Department of Molecular Biosciences, University of Oslo, Norway2005 7 11 2005 4 39 39 17 2 2005 7 11 2005 Copyright © 2005 Kresse et al; licensee BioMed Central Ltd.2005Kresse et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Amplification of the q21-q23 region on chromosome 1 is frequently found in sarcomas and a variety of other solid tumours. Previous analyses of sarcomas have indicated the presence of at least two separate amplicons within this region, one located in 1q21 and one located near the apolipoprotein A-II (APOA2) gene in 1q23. In this study we have mapped and characterized the amplicon in 1q23 in more detail.
Results
We have used fluorescence in situ hybridisation (FISH) and microarray-based comparative genomic hybridisation (array CGH) to map and define the borders of the amplicon in 10 sarcomas. A subregion of approximately 800 kb was identified as the core of the amplicon. The amplification patterns of nine possible candidate target genes located to this subregion were determined by Southern blot analysis. The genes activating transcription factor 6 (ATF6) and dual specificity phosphatase 12 (DUSP12) showed the highest level of amplification, and they were also shown to be over-expressed by quantitative real-time reverse transcription PCR (RT-PCR). In general, the level of expression reflected the level of amplification in the different tumours. DUSP12 was expressed significantly higher than ATF6 in a subset of the tumours. In addition, two genes known to be transcriptionally activated by ATF6, glucose-regulated protein 78 kDa and -94 kDa (GRP78 and GRP94), were shown to be over-expressed in the tumours that showed over-expression of ATF6.
Conclusion
ATF6 and DUSP12 seem to be the most likely candidate target genes for the 1q23 amplification in sarcomas. Both genes have possible roles in promoting cell growth, which makes them interesting candidate targets.
==== Body
Background
Gains or amplification of the long arm of chromosome 1 is among the most common chromosomal abnormalities in human neoplasia [1]. Local gains or high-level amplification affecting 1q21-q23 is particularly frequent, and was first described for sarcomas [2-4]. Sarcomas are a heterogeneous group of malignant tumours of various supporting- and connective tissues, ranging from the almost benign well-differentiated liposarcomas (WDLS) to aggressive tumour forms, such as fibrosarcomas, osteosarcomas and malignant fibrous histiocytomas (MFH) [5]. Although infrequent, these tumours have been widely studied at the molecular level, and this has provided insight into mechanisms of importance for tumour development in general. Notably, amplification and over-expression of MDM2 and CDK4, representing alternative pathways for inactivation of the tumour suppressors p53 and pRb, respectively, were first described for this group of tumours [6,7].
A variety of solid tumours show amplification of 1q21-q22, for instance 70–80 % of hepatocellular carcinomas and 25–30 % of ovarian cancers [8,9]. Studies of renal clear cell, hepatocellular and colorectal carcinomas revealed a higher frequency of 1q21-q23 gains in metastatic tumours [10,11], and gains of 1q21-q25 showed a trend toward short overall survival in high-grade osteosarcomas and neuroblastomas [12,13]. However, also sarcomas of borderline malignancy, such as WDLS, show frequent amplification of 1q21-q23 [2,3,14], and 1q21 is recurrently gained in desmoid tumours [15].
A more detailed molecular analysis of sarcomas has indicated the presence of at least two separate amplicons within 1q21-q23, located in 1q21 and 1q23, respectively [16]. The amplicon in 1q21 was most frequent, and was represented by a yeast artificial chromosome (YAC) clone, 789f2, located proximal to the S100-genes. The YAC detected high amplification levels, and was subsequently used to clone three novel candidate target genes that are highly amplified and over-expressed in both sarcomas and breast cancer [17]. The 1q23 amplicon was first observed in one single tumour (LS21), where a probe for the apolipoprotein A-II (APOA2) gene detected high amplification levels [16]. Furthermore, the extent of the amplified region around 1q21 as observed in comparative genomic hybridisation (CGH) analyses was variable, covering only 1q21 in some tumours and 1q21-q23 or -q25 in other tumours [3,4]. These observations indicate that also amplified genes located more distally of 1q21 may be of importance in sarcoma development or progression.
We have now used a bacterial artificial chromosome (BAC) contig of approximately 7 Mb, according to Ensembl (, assembly December 14, 2004) [18] and the UCSC Genome Browser (, assembly May 2004) [19], to map and characterize the amplified region around APOA2 in LS21 using fluorescence in situ hybridisation (FISH). The two BACs that showed particularly high copy numbers were further analysed in nine additional sarcomas known to have gains of the 1q21-q23 region [3,16] (and unpublished results). Amplicon mapping was also done by microarray-based CGH (array CGH). The amplification patterns of nine possible candidate target genes were determined by Southern blot analysis, and quantitative real-time reverse transcription PCR (RT-PCR) was used to determine the expression levels of the two candidate target genes that showed the highest level of amplification. In addition, the expression level of two genes transcriptionally activated by one of the candidate target gene was analysed.
Results
Mapping of the 1q23 amplicon by FISH
Fifty-nine overlapping BACs, covering approximately 7 Mb in the q23 region of chromosome 1, were hybridised to interphase nuclei from liposarcoma LS21. Figure 1A shows the copy numbers detected by all the BACs.
Figure 1 DNA copy number by FISH. (A) FISH analysis with 59 BACs on interphase nuclei from liposarcoma LS21. The BACs are listed from the centromeric to the telomeric side. The colour of the shading indicates the range of signals observed per nucleus, and the area the percentage of nuclei within each group of signals. The localisation of APOA2 in BAC RP11-297K8 is indicated, as well as the distance position of some of the BACs (the intervals are not equal since the BACs are not distributed evenly throughout the region). The two BACs used further in Figure 1B are highlighted, and the BACs used for array CGH in Figure 2 are marked with an asterisk (*). There is a gap of approximately 200 kb between BACs RP11-297K8 and RP11-5K23. After these analyses were performed, the gap was covered by the BAC RP11-122G18. Since no more material is available from LS21, this BAC has not been tested. (B) FISH analysis with BACs RP11-110D4 and RP11-195G14 on interphase nuclei from 10 sarcomas (LMS, leiomyosarcoma; LS, liposarcoma; MFH, malignant fibrous histiocytoma; MS, malignant schwannoma, suffix 'x', xenograft). (C) FISH analysis with four BACs proximal and three distal of RP11-110D4 and RP11-195G14 on interphase nuclei from LS3x, LS21 and MS8x.
Copy numbers varied throughout the region, but a subregion defined by 12 overlapping BACs, RP11-312J18 through RP11-565P22, was highly amplified in this tumour compared to the flanking areas. Furthermore, two BACs within this amplified unit, RP11-110D4 and RP11-195G14, detected higher copy numbers than any of the other BACs tested. These two BACs are located approximately 400 kb distal to the APOA2 gene that originally identified this amplicon. APOA2 is located in BAC RP11-297K8 according to Ensembl. The localisation of BACs RP11-110D4 and RP11-195G14 to 1q23 was confirmed by hybridisation to metaphase nuclei from normal leukocytes (not shown).
To further analyse the amplification frequency of the sequences represented by these two BACs, we determined the copy numbers in nine additional sarcomas known to have gains of 1q21-q23 [3,16], and unpublished results]. The results are presented in Figure 1B. In all cases, amplification was detected in a substantial fraction of the nuclei, especially for LS3x and MS8x. Only two tumours, LMS2x and LS6, showed normal copy numbers in more than 60 % of the nuclei.
The borders of the amplicon were in addition defined for LS3x and MS8x, where high-level amplification was found in 20–50 % of the nuclei. Four BACs located proximal and three located distal to the two BACs detecting the highest copy number were hybridised to interphase nuclei (Figure 1C). The amplicon in LS3x was represented by a few more BACs than the amplicon in LS21 and MS8x. For all three tumours, BACs RP11-110D4 and RP11-195G14 detected the highest level of amplification.
Mapping of the 1q23 amplicon by array CGH
Mapping of the 1q23 amplicon was also done by array CGH. The 10 tumours analysed by FISH were hybridised to a genomic microarray covering the minimal tiling-path of the 1q12-q25 region. The results from 15 BACs, covering the region in 1q23 that showed highest amplification level by FISH (Figure 1A), are presented here. The dataset with these 15 BACs can be viewed in the microarray database ArrayExpress (, accession number E-MEXP-427) [20].
Figure 2 shows the relative copy numbers compared to a normal reference of the overlapping BACs. In four of the tumours, amplification levels above 3-fold were observed, whereas the relative copy number of all the BACs was lower in the other six tumours. For MS8x, the subregion covered by BACs RP11-5K23 through RP11-384L19 showed the highest copy number, with RP11-195G14 and RP11-456P18 detecting relative copy numbers of more than 14. For LS6, the highest copy number was detected by RP11-227F8. LS21 showed relative copy numbers of 5–6 of the region covered by RP11-5K23 through RP11-456P18, but a higher level of the region covered by BACs RP11-312J18 through RP11-297K8, with RP11-137A12 and RP11-297K8 detecting a relative copy number of 7. These two BACs detected the highest copy number also in LS43.
Figure 2 DNA copy number by array CGH. Relative DNA copy number of the region represented by BACs RP11-404F10 through RP11-565P22 in 10 sarcomas. The localisation of BACs RP11-381D2, RP11-110D4, RP11-195G14 and RP11-336H14 is based on previous contig maps from the Wellcome Trust Sanger Institute (dotted line), since these are removed from the minimal tiling-path in Ensembl. *BAC RP11-122G18 is not present in the array.
The array CGH results showed in addition other clusters of BACs detecting high copy numbers (data not shown), indicating the presence of several separate amplicons within the 1q12-q25 region, which will be described in a later publication.
Identification of candidate target genes
Ensembl and the UCSC Genome Browser were used to search for possible candidate target genes. Based on sequence information, these programs visually display possible open reading frames, as well as predicted and known genes.
BACs RP11-110D4 and RP11-195G14, used here for FISH and array CGH, have later been omitted from the minimal tiling-path (Ensembl). The overlapping BACs RP11-5K23, RP11-474I16 and RP11-456P18, which constitute the minimal tiling-path, now represent the corresponding region.
Within the approximately 800 kb region covered by the BACs RP11-5K23 through RP11-384L19, 11 genes were mapped according to Ensembl. The genes are FCGR2A, -2B, and -3A, a family of low affinity immunoglobulin gamma FC receptors; HSPA6, heat shock 70 kDa protein 6; DUSP12, dual specificity phosphatase 12; ATF6, activating transcription factor 6; OLFML2B, olfactomedin-like 2B; three genes referred to by RefSeqID: NP_116127, FC receptor homolog expressed in B cells (FREB); NP_055512, C-terminal PDZ domain ligand of neuronal nitric oxide synthase (CAPON) and NP_001002901, with no description; and one "novel" gene.
The UCSC Genome Browser indicated the same genes as Ensembl, and in addition FCGR3B, another member of the low affinity immunoglobulin gamma FC receptors family.
Southern blot analysis of the candidate target genes
Copy numbers of nine of the candidate target genes were further analysed by Southern blot analysis. The relative copy numbers of the genes compared to APOB are presented in Figure 3. The candidate target genes that were not tested were one "novel" gene, one gene with RefSeqID (both with no description), and OLFML2B, which appeared in Ensembl after the analyses were completed.
Figure 3 Gene copy number by Southern blot analysis. Relative copy number of nine of the candidate target genes normalized with APOB in 10 sarcomas. The localisation of BACs RP11-110D4 and RP11-195G14 is based on previous contig maps from the Wellcome Trust Sanger Institute (dotted line), since these are removed from the minimal tiling-path in Ensembl. Gene localisations are based on Ensembl and the UCSC Genome Browser, with the genes listed from the centromeric to the telomeric side. Of the genes located in BACs RP11-297K8 and RP11-122G18, only APOA2 is shown. The densitometrically determined levels of amplification are divided into five categories as indicated.
For tumours LS3x, LS21, LS43 and MS8x, the genes ATF6 and DUSP12 showed the highest amplification level in general, 3–5 fold in LS3x and 5–10 fold in LS21 and LS43 (except ATF6), and particularly high copy numbers (>10 fold) in MS8x. FREB also showed high amplification levels, except in MS8x where the level was 5-fold (approximately three times less than ATF6 and DUSP12). For the other six tumours, the amplification level of all the genes was either normal or 2–3 fold.
APOA2 showed the same amplification levels as identified previously [16], except that the copy number increase was determined to be 5–10 fold instead of >10 fold in LS21, and 2–3 fold instead of <2 fold in LMS15, either because of experimental variation or because different parts of the tumours were analysed. APOA2 is located in BAC RP11-297K8 (Ensembl). By FISH analysis, RP11-297K8 detected high-level amplification in LS21 (32 % of the nuclei), LS3x (13 % of the nuclei) and MS8x (16 % of the nuclei) (Figure 1C), as one would expect.
Expression analysis of the candidate target genes by quantitative real-time RT-PCR
Quantitative real-time RT-PCR was used to determine the expression levels of the two genes ATF6 and DUSP12, which showed the highest level of amplification by Southern blot analysis. In addition, the expression level of two genes known to be transcriptionally activated by ATF6 was analysed, GRP78 and GRP94, glucose-regulated protein 78 kDa and -94 kDa. Figure 4 shows the expression level of ATF6, DUSP12, GRP78 and GRP94 compared to the average expression of the endogenous control genes β-2-microglobulin (B2M), glyceraldehyde-3-phosphate dehydrogenase (GAPDH) and TATA box binding protein (TBP). The average relative expression level of the tumours with normal copy number of ATF6 and DUSP12 (LS2, LS6 and MFH36) has been set to 1.
Figure 4 Gene expression level by quantitative real-time RT-PCR. Relative expression level of the genes ATF6, DUSP12, GRP78 and GRP94 in 10 sarcomas. The expression levels have been normalized with the average expression of the endogenous control genes B2M, GAPDH and TBP. The average relative expression level of the tumours with normal copy number of ATF6 and DUSP12 (LS2, LS6 and MFH36) has been set to 1. The relative copy number of ATF6 and DUSP12 determined by Southern blot analysis (Figure 3) is also shown.
LS21 and MS8x showed the highest level of expression of both candidate target genes, ATF6 was over-expressed 8–9 fold whereas DUSP12 was over-expressed 11-fold in LS21 and 25-fold in MS8x. LS3x showed 3-fold over-expression of both genes, and for LMS2x and MFH19, DUSP12 was over-expressed 2–3 fold whereas ATF6 was expressed at the same level as the three tumours with normal copy number. LMS15 and LS43 showed the same expression levels of both ATF6 and DUSP12 as the tumours with normal copy number of the genes.
DUSP12 was significantly higher expressed than ATF6 in LMS2x, MFH19, MS8x (all p < 0,001), LS21 and LS43 (both p < 0,01), whereas ATF6 was significantly higher expressed than DUSP12 in LMS15 (p < 0,01). There was no significant difference in expression of the two genes in the rest of the tumours.
GRP78 and GRP94 were over-expressed in the same tumours that showed over-expression of ATF6. GRP78 was over-expressed 8-fold in MS8x, 5-fold in LS21 and 3-fold in LS3x, whereas GRP94 was over-expressed 3-fold in LS3x and 2-fold in LS21 and MS8x. In addition, MFH19 showed a 2-fold increase of both genes, whereas LMS2x showed a 2-fold increase for only GRP78. LMS15 and LS43 showed the same expression level of both GRP78 and GRP94 as the tumours with normal copy number of ATF6 and DUSP12, as they did for ATF6 and DUSP12.
Discussion
CGH analyses of sarcomas show that the amplified part of 1q is variable. In some tumours, only 1q21 is highly amplified, whereas other cases have amplification of the whole 1q21-q23 or 1q21-q25 region [3,4]. These observations suggest the presence of multiple, separate amplicons each containing target genes expected to be important for the development or progression of these tumours.
In this study, we did FISH with a contig of BACs spanning 7 Mb around APOA2 to map and characterize this amplicon in more detail in liposarcoma LS21, which showed the highest copy number of APOA2 in previous analyses [16]. Our results showed that the core of the amplicon could be defined by 12 partly overlapping BACs (Figure 1A). The two BAC clones that detected the highest amplification levels, RP11-110D4 and RP11-195G14, were located approximately 400 kb distally of APOA2. Partial mapping of the amplicon in two other tumours confirmed that the most amplified part was covered by these two BACs (Figure 1C), which detected moderate to high-level amplification in eight of the 10 sarcomas analysed (Figure 1B).
The 1q21 amplicon, as represented by YAC 789f2, was present in all tumours with 1q21-q23 amplification analysed here [3,16] (and unpublished results), and it was generally present at higher copy numbers than the amplicon near APOA2 in 1q23. This pattern seemed neither to be dependent on the sarcoma subtype, nor on the aggressiveness of the tumour. Both amplicons were for instance observed in WDLS (LS2 and LS21) as well as in the more aggressive leiomyosarcomas. However, the results presented here indicate that target genes located in 1q23 may be important for tumour development or progression in a subset of sarcomas, as this clearly is a separate amplicon.
Amplicon mapping was confirmed by array CGH using tiling-path probes for this region (Figure 2). For the most amplified subregion identified by FISH, four of the tumours showed a variable amplification pattern, whereas the other six tumours showed a relative copy number between 1 and 3. The approximately 800 kb region covered by the BACs RP11-5K23 through RP11-384L19 appeared to be the core of the amplicon, where LS6, LS21 and especially MS8x showed high amplification levels. But for LS21 and LS43, the region covered by the BACs RP11-312J18 through RP11-5K23 also showed high levels, suggesting that additional genes may be important for these two cases.
The results obtained by FISH and array CGH were in general consistent, but differences were also observed, in particular for tumours LS3x and LS43. FISH detected high copy numbers in a substantial fraction of the cells in LS3x, whereas array CGH detected a relative copy number around 3. For LS43 it was the other way around, with higher copy numbers detected by array CGH than FISH. Different pieces of the tumour were used to make interphase nuclei for FISH and isolate DNA for array CGH, and the observed discrepancy could thus be explained by tumour heterogeneity. Heterogeneous amplification status has been detected previously for some tumours, like LS21 [16]. Also, normal cells in the tumours would cause amplification levels of the cancer cells to be underestimated by array CGH, and a subpopulation of cells with high amplification levels may go undetected. FISH analysis, although very laborious, would however detect such a subpopulation.
In order to identify possible candidate target genes we used Ensembl and the UCSC Genome Browser, and Southern blot analysis was performed to determine the copy numbers of nine of the candidate target genes (Figure 3). For tumours LS3x, LS21, LS43 and MS8x, the genes ATF6 and DUSP12 showed the highest amplification level in general, suggesting that one of them may be the target for this amplification. Both genes were located within the two BACs that detected the highest level of amplification.
Dual specificity phosphatase 12 (DUSP12) is a member of the VH1-like dual specificity subfamily of protein phosphatases, which may dephosphorylate both phosphoserine/threonine and phosphotyrosine residues [21]. This gene is also termed YVH1, being the Homo sapiens ortholog of the Saccharomyces cerevisiae gene YVH1 protein-tyrosine phosphatase. Little is known about the human DUSP12/YVH1 protein function, but S. cerevisiae YVH1 is involved in cell growth, meiosis and sporulation [22], and inactivation of the S. cerevisiae YVH1 gene results in a striking increase in yeast doubling time [23]. A similar effect has been observed in the opportunistic fungal pathogen Candida albicans, where YVH1 also contributes to pathogenicity [24]. Interestingly, the human YVH1 gene has been shown to rescue the slow growth defect in yeast caused by disruption of the S. cerevisiae YVH1 gene [25]. Since DUSP12 may be involved in cell proliferation, it is possible that amplification of DUSP12 stimulates tumour growth.
In addition, S. cerevisiae YVH1 has been shown to interact with the yeast pescadillo homolog (YPH1) [26], and this interaction has also been observed in the malaria parasite Plasmodium falciparum with the orthologs of YVH1 and pescadillo [27]. Pescadillo is essential for ribosome biogenesis, nucleogenesis and mammalian cell proliferation [28,29], and disruption to its function results in cell cycle arrest [29]. An increased expression of pescadillo protein has been demonstrated in malignant cells [28], implying that pescadillo may contribute toward tumour progression. Thus, the interaction between YVH1 and YPH1 also connects DUSP12/YVH1 to cell proliferation and cell cycle regulation.
Activating transcription factor 6 (ATF6) is a member of the basic-leucine zipper (bZIP) family of transcription factors. It is involved in the endoplasmic reticulum (ER) stress response pathway, and activates expression of genes induced by the ER stress response [30] (and references therein). Interestingly, the ER stress response is already implicated in sarcoma biology, as the CHOP/GADD153 gene, involved in these processes [31], is translocated in myxoid liposarcomas [32] and also sometimes amplified in sarcomas [33].
ATF6 has been shown to activate the promoters of the genes glucose-regulated protein 78 kDa and -94 kDa (GRP78 and GRP94) [34,35], which function as ER chaperones. GRP78 is also termed immunoglobulin heavy chain binding protein (BiP) and heat shock 70kDa protein 5 (HSPA5), while GRP94 is also termed tumour rejection (gp96) antigen 1 (TRA1).
GRP78 and GRP94 have been shown to protect cells against apoptosis [36,37] (and references therein), and this anti-apoptotic function suggests that induction of these genes could lead to cancer progression and drug resistance in neoplastic cells. Several studies have correlated over-expression of GRP78 and GRP94 with tumour growth [36,37] (and references therein), most likely because of inhibition of apoptosis. Thus, it is possible that amplification of ATF6 could lead to over-expression of GRP78 and GRP94, thereby giving the cells a growth advantage or resistance to chemotherapy.
The expression levels of ATF6 and DUSP12 were determined by quantitative real-time RT-PCR (Figure 4), and in general the expression level of ATF6 and DUSP12 reflected the amplification level in the different tumours. LS21 and MS8x showed the highest level of expression, with DUSP12 being as high as 25-fold over-expressed in MS8x. DUSP12 was expressed significantly higher than ATF6 in these two tumours (p < 0,001 for MS8x and p < 0,01 for LS21), and also in LMS2x and MFH19 (both p < 0,001). Thus, except for LS3x, DUSP12 is significantly higher expressed than ATF6 in all the tumours that showed at least 2-fold over-expression of either or both genes, suggesting that DUSP12 may be the real target for the amplification.
In order to investigate whether amplified ATF6 is active, the expression levels of GRP78 and GRP94 were determined (Figure 4). Interestingly, both genes were also over-expressed in the tumours that showed high expression of ATF6 (LS3x, LS21 and MS8x). GRP78 showed especially high levels, being over-expressed 8–9 fold in LS21 and MS8x. In addition, expression of both genes was increased 2-fold in MFH19 and also in LMS2x for GRP78, two tumours with similar expression level of ATF6 as the tumours with normal copy number of ATF6. However, the expression level of GRP78 and GRP94 in these two tumours is lower than what is observed in the tumours with over-expression of ATF6, and may possibly be caused by other regulatory mechanisms. Thus, amplification and over-expression of ATF6 seem to cause over-expression of particular GRP78 but also GRP94, thereby implying that amplification of ATF6 has a functional role.
Conclusion
Based on their consistent association with this amplicon and possible roles in promoting cell growth, both ATF6 and DUSP12 represent interesting candidate target genes for the 1q23 amplification in sarcomas. DUSP12 seems to be the most likely target based on the significantly higher expression in a subset of the tumours, but since relevant genes activated by ATF6 also are highly expressed in tumours with over-expression of ATF6, both genes are considered potential targets. Further functional analyses are required to determine the role of these proteins in sarcoma development or progression.
Methods
Specimens
Ten sarcomas with known alterations of 1q21-q23 were analysed in this study, five liposarcomas (LS), two leiomyosarcomas (LMS), two malignant fibrous histiocytomas (MFH) and one malignant peripheral nerve sheath tumour (MPNST) (previously termed malignant schwannoma, MS). Histopathological and clinical characteristics of these tumours have been previously described [16,17].
Fluorescence in situ hybridisation (FISH)
Fifty-nine overlapping BACs were used as probes for FISH. All the BACs were from the clone library RPCI-11 [38], kindly provided by the Wellcome Trust Sanger Institute . The precise localisation of each BAC was based on sequence alignment by basic local alignment search tool (BLAST), performed by the Wellcome Trust Sanger Institute.
BAC DNA was isolated by standard methods and labeled with biotin-16-dUTP or digoxygenin-11-dUTP (Roche) by use of the BioNick Labeling System (Invitrogen). For each hybridisation, 300 ng of labeled DNA was ethanol precipitated together with 10 μg human Cot-1 DNA (Invitrogen). The precipitated DNA was dissolved in 15 μl hybridisation buffer (50 % formamide, 10 % dextran sulphate, 2 × SSC).
Preparation of interphase nuclei from tumour tissue was done as previously described [16]. The slides were thawed and immersed in 70 % ethanol for at least 1 hour, air-dried and treated with 0,4 mg/ml pepsin for 20 minutes at 37°C. After this, the slides were washed three times for 5 minutes in 1 × PBS, fixated in 1 % formaldehyde/1 × PBS for 10 minutes at room temperature and washed three times for 5 minutes in 1 × PBS. The slides were dehydrated in ethanol (70, 90, 96 and 100 %) and air-dried.
The slides were denatured for 2 minutes at 74°C in 70 % de-ionized formamide/2 × SSC, immediately transferred to ice-cold 70 % ethanol, further dehydrated and air-dried. The labeled DNA was denatured for 10 minutes at 80°C, prehybridised for at least 30 minutes at 37°C and applied to slides at room temperature. Hybridisation was performed overnight at 37°C. After hybridisation, slides were washed three times for 10 minutes in 50 % formamide/2 × SSC at 45°C, and three times for 10 minutes in 0,1 × SSC at 45°C.
Signals from biotin-labelled probes were detected with avidin-Cy3 (Amersham Biosciences), and for detection of digoxygenin-labelled probes we used a FITC-labelled sheep-anti-digoxygenin antibody followed by ALEXA 488-labelled donkey-anti-sheep (Molecular Probes). For probes that gave weak signals, signals were amplified by deposition of FITC- or biotin-labelled thyramides, essentially as described in the protocols supplied by the manufacturer (TSA-direct and -indirect, NEN™ Life Science Products). Signals from biotinylated thyramides were detected with FITC-labelled streptavidin (NEN). The interphase nuclei were counterstained with 4',6-diamino-2-phenylindole (DAPI) and mounted in anti-fade solution (Vector Laboratories).
Hybridised slides were examined visually using a Zeiss Axioplan microscope equipped with appropriate filters for excitation of DAPI, DAPI/FITC or DAPI/Rhodamine (Cy3). The slides were manually scanned at 63 × or 100 × magnification with DAPI excitation to localise the interphases. For each probe, the number of spots was counted in at least 100 nuclei. Amplification levels were grouped into three categories; normal (two signals), moderate (3–9 signals) and high (10 or more signals).
Microarray-based comparative genomic hybridisation (array CGH)
Genomic microarrays covering the 1q12-q25 region were made using overlapping BACs and P1 artificial chromosomes (PACs). The BACs were from the clone library RPCI-11 [38], whereas the PACs were from the clone libraries RPCI-1, -3, -4 and -5. All clones were kindly provided by the Wellcome Trust Sanger Institute . The precise localisation of each BAC and PAC was based on sequence alignment by BLAST, performed by the Wellcome Trust Sanger Institute.
The results from 15 of the BACs, covering the region in 1q23 that showed highest amplification level by FISH, are presented here. The dataset with these 15 BACs has been submitted to ArrayExpress, the microarray database of the European Bioinformatics Institute (, accession number E-MEXP-427) [20].
Isolation of BAC and PAC DNA, amplification by DOP-PCR and preparation of microarrays were done as previously described [39]. The PCR products were arrayed in quadruplicate onto amine-binding slides (CodeLink, Amersham Biosciences) using a MicroGrid II arrayer (BioRobotics).
Genomic DNA from tumour tissues was isolated by standard methods as described previously [40]. Normal female or male DNA (Promega) was used as a reference. Labeling of the DNA was done as previously described [39], with a few modifications. Here, 500 ng total genomic DNA was labeled using 1,5 μl 1 mM Cy3-dCTP or Cy5-dCTP (Amersham Biosciences) in a total volume of 100 μl.
The labeled tumour and reference DNA were combined and ethanol precipitated together with 135 μg human Cot-1 DNA (Invitrogen). The precipitated DNA was dissolved in 108 μl hybridisation buffer (50 % formamide, 10 % dextran sulphate, 4 % SDS, 2 × SSC) and 4 μl 100 mg/ml yeast tRNA (Invitrogen). The DNA was denatured for 10 minutes at 70°C and prehybridised for at least 30 minutes at 37°C.
Hybridisation was performed using an automated hybridisation station, GeneTAC (Genomic Solutions/Perkin Elmer), agitating the hybridisation solution for 48 hours at 37°C. After hybridisation, the slides were washed with 50 % formamide/2 × SSC at 48°C, 2 × SSC/0,1 % SDS at 48°C and PN-buffer (0,1 M NaH2PO4 plus 0,1 M Na2HPO4 to obtain pH 8/0,1 % NP-40) at 25°C. For all three solutions, the hybridisation station washed 5 cycles with a flow time of 20 seconds and a hold time of 40 seconds for each cycle. After removal from the hybridisation station, the slides were rinsed briefly in 0,05 × SSC and dried by spinning in a centrifuge.
The arrays were scanned by use of an Agilent G2565BA scanner (Agilent Technologies). The acquired microarray images were analysed using GenePix Pro 4.1 (Axon Laboratories). The spots were automatically segmented and manually adjusted where necessary. The fluorescent intensities and the local background of the two dyes were calculated for each spot. Further data processing, including filtering and normalisation, was done using M-CGH, a MATLAB toolbox specifically designed for this purpose [41].
Southern blot analysis
Eleven human cDNA clones were used as probes for Southern blot analysis. The clones for the nine candidate target genes were from the I.M.A.G.E. Consortium [LLNL] cDNA clones library [42], provided by Research Genetics. The I.M.A.G.E. Consortium CloneID and GenBank accession numbers of the clones are as follows: ATF6 (417251, W87752); CAPON (1860405, AI198232); DUSP12 (843328, AA485951); FCGR2A (868380, AA634109); FCGR2B (138369, R68106); FCGR3A (450155, AA703460); FCGR3B (51447, H20822); FREB (291871, W02963) and HSPA6 (2310335, AI654494).
In addition, a cDNA for APOA2 [43] was used. As a control probe, the apolipoprotein B (APOB) gene, located in the p24 region of chromosome 2, was used [44]. Previous work has shown no amplification of this chromosomal region in neither our sarcoma panel [3,4,45] nor liposarcomas from other groups [2,46]. All cDNA clones were sequence verified.
DNA extraction from tumour tissues, digestion, preparation of filter blots and hybridisation were done as previously described [40]. Quantitation of signal intensity was done by two-dimensional densitometry on a Molecular Dynamics laser densitometer. To correct for unequal sample loading, the gene-specific signals were calibrated to the relative signals obtained from the APOB control probe. Amplification levels were calculated by comparison with signals from normal controls (leukocytes), and grouped into five categories; <2 (normal), 2–3, 3–5, 5–10 and >10.
Quantitative real-time reverse transcription PCR (RT-PCR)
Quantitative real-time RT-PCR was performed using TaqMan® Gene Expression Assays (Applied Biosystems). The expression level was determined for the candidate target genes ATF6 (assay ID Hs00232586_m1) and DUSP12 (assay ID Hs00170898_m1). In addition, the expression level of GRP78 (also termed HSPA5, assay ID Hs00946088_g1) and GRP94 (also termed TRA1, assay ID Hs00427665_g1) was analysed. The genes B2M (assay ID Hs99999907_m1), GAPDH (assay ID Hs99999905_m1) and TBP (assay ID Hs99999910_m1) were used as endogenous controls for normalization. These housekeeping genes were chosen since they showed low variability by microarray expression profiling of our panel of sarcomas (Namløs, Berner, Myklebost et al., unpublished).
Frozen tumour tissue was pulverized in liquid nitrogen, and total RNA was extracted using Trizol (Invitrogen) according to the manufacturer's instructions. Universal Human Reference RNA (Stratagene) was used as a reference. cDNA synthesis was performed using the High-Capacity cDNA Archive Kit, essentially as described in the protocols supplied by the manufacturer (Applied Biosystems).
The PCR amplification was performed according to the manufacturer's instructions using the ABI Prism 7000 Sequence Detection System (Applied Biosystems). The cycling conditions comprised 10 minutes polymerase activation at 95°C and 40 cycles of 95°C for 15 seconds and 60°C for 1 minute. Each assay included (in triplicate): a standard curve of four serial dilutions of the Universal Human Reference RNA cDNA (ranging from 50 ng to 50 pg), a no-template control and 2 ng of each tumour cDNA.
Gene expression in the tumours was determined from the standard curves, and the expression level of ATF6, DUSP12, GRP78 and GRP94 was normalized with the average expression of the three endogenous controls.
Authors' contributions
SHK performed the experiments and analyses and drafted the manuscript. JMB participated in performing the Southern blot experiments and analyses. LAMZ participated in performing the array CGH experiments and analyses. SGG provided genomic clones and information. WLK and JWG participated in establishing the array CGH technology. AF and OM conceived of the study, participated in its design and coordination and helped to draft the manuscript.
Acknowledgements
We thank Erik B. Paulsen and Magne Skårn for excellent technical assistance, and the members of the Wellcome Trust Sanger Institute chromosome 1 mapping and sequencing teams for provision of genomic clones and information. The genomic microarrays were provided by the Norwegian Microarray Consortium (NMC) at the national technology platform, and supported by the functional genomics programme (FUGE) in the Research Council of Norway. This work was supported by the University of Oslo (EMBIO), the Norwegian Cancer Society and the Norwegian Foundation for Health and Rehabilitation.
==== Refs
Schutte BC Carpten JD Forus A Gregory SG Horii A White PS Report and abstracts of the sixth international workshop on human chromosome 1 mapping 2000. Cytogenet Cell Genet 2001 92 23 48 11306795 10.1159/000056867
Szymanska J Tarkkanen M Wiklund T Virolainen M Blomqvist C Asko-Seljavaara S Tukiainen E Elomaa I Knuutila S Gains and losses of DNA sequences in liposarcomas evaluated by comparative genomic hybridization Genes Chromosomes Cancer 1996 15 89 94 8834171 10.1002/(SICI)1098-2264(199602)15:2<89::AID-GCC2>3.0.CO;2-#
Forus A Weghuis DO Smeets D Fodstad O Myklebost O van Kessel AG Comparative genomic hybridization analysis of human sarcomas: I. Occurrence of genomic imbalances and identification of a novel major amplicon at 1q21-q22 in soft tissue sarcomas Genes Chromosomes Cancer 1995 14 8 14 8527390
Forus A Weghuis DO Smeets D Fodstad O Myklebost O van Kessel AG Comparative genomic hybridization analysis of human sarcomas: II. Identification of novel amplicons at 6p and 17p in osteosarcomas Genes Chromosomes Cancer 1995 14 15 21 8527379
Enzinger FM Weiss SW Soft tissue tumors 1995 3rd St. Louis, Mosby
Oliner JD Kinzler KW Meltzer PS George DL Vogelstein B Amplification of a gene encoding a p53-associated protein in human sarcomas Nature 1992 358 80 83 1614537 10.1038/358080a0
Khatib ZA Matsushime H Valentine M Shapiro DN Sherr CJ Look AT Coamplification of the CDK4 gene with MDM2 and GLI in human sarcomas Cancer Res 1993 53 5535 5541 8221695
Wong N Lai P Lee SW Fan S Pang E Liew CT Sheng Z Lau JW Johnson PJ Assessment of genetic changes in hepatocellular carcinoma by comparative genomic hybridization analysis: relationship to disease stage, tumor size, and cirrhosis Am J Pathol 1999 154 37 43 9916916
Zimonjic DB Keck CL Thorgeirsson SS Popescu NC Novel recurrent genetic imbalances in human hepatocellular carcinoma cell lines identified by comparative genomic hybridization Hepatology 1999 29 1208 1214 10094966 10.1002/hep.510290410
Gronwald J Storkel S Holtgreve-Grez H Hadaczek P Brinkschmidt C Jauch A Lubinski J Cremer T Comparison of DNA gains and losses in primary renal clear cell carcinomas and metastatic sites: importance of 1q and 3p copy number changes in metastatic events Cancer Res 1997 57 481 487 9012478
Qin LX Tang ZY Sham JS Ma ZC Ye SL Zhou XD Wu ZQ Trent JM Guan XY The association of chromosome 8p deletion and tumor metastasis in human hepatocellular carcinoma Cancer Res 1999 59 5662 5665 10582679
Tarkkanen M Elomaa I Blomqvist C Kivioja AH Kellokumpu-Lehtinen P Bohling T Valle J Knuutila S DNA sequence copy number increase at 8q: a potential new prognostic marker in high-grade osteosarcoma Int J Cancer 1999 84 114 121 10096241 10.1002/(SICI)1097-0215(19990420)84:2<114::AID-IJC4>3.0.CO;2-Q
Hirai M Yoshida S Kashiwagi H Kawamura T Ishikawa T Kaneko M Ohkawa H Nakagawara A Miwa M Uchida K 1q23 gain is associated with progressive neuroblastoma resistant to aggressive treatment Genes Chromosomes Cancer 1999 25 261 269 10379872 10.1002/(SICI)1098-2264(199907)25:3<261::AID-GCC8>3.0.CO;2-#
Pedeutour F Forus A Coindre JM Berner JM Nicolo G Michiels JF Terrier P Ranchere-Vince D Collin F Myklebost O Turc-Carel C Structure of the supernumerary ring and giant rod chromosomes in adipose tissue tumors Genes Chromosomes Cancer 1999 24 30 41 9892106 10.1002/(SICI)1098-2264(199901)24:1<30::AID-GCC5>3.0.CO;2-P
Larramendy ML Virolainen M Tukiainen E Elomaa I Knuutila S Chromosome band 1q21 is recurrently gained in desmoid tumors Genes Chromosomes Cancer 1998 23 183 186 9739022 10.1002/(SICI)1098-2264(199810)23:2<183::AID-GCC12>3.0.CO;2-P
Forus A Berner JM Meza-Zepeda LA Saeter G Mischke D Fodstad O Myklebost O Molecular characterization of a novel amplicon at 1q21-q22 frequently observed in human sarcomas Br J Cancer 1998 78 495 503 9716033
Meza-Zepeda LA Forus A Lygren B Dahlberg AB Godager LH South AP Marenholz I Lioumi M Florenes VA Maelandsmo GM Serra M Mischke D Nizetic D Ragoussis J Tarkkanen M Nesland JM Knuutila S Myklebost O Positional cloning identifies a novel cyclophilin as a candidate amplified oncogene in 1q21 Oncogene 2002 21 2261 2269 11948409 10.1038/sj.onc.1205339
Hubbard T Barker D Birney E Cameron G Chen Y Clark L Cox T Cuff J Curwen V Down T Durbin R Eyras E Gilbert J Hammond M Huminiecki L Kasprzyk A Lehvaslaiho H Lijnzaad P Melsopp C Mongin E Pettett R Pocock M Potter S Rust A Schmidt E Searle S Slater G Smith J Spooner W Stabenau A Stalker J Stupka E Ureta-Vidal A Vastrik I Clamp M The Ensembl genome database project Nucleic Acids Res 2002 30 38 41 11752248 10.1093/nar/30.1.38
Kent WJ Sugnet CW Furey TS Roskin KM Pringle TH Zahler AM Haussler D The human genome browser at UCSC Genome Res 2002 12 996 1006 12045153 10.1101/gr.229102. Article published online before print in May 2002
Brazma A Parkinson H Sarkans U Shojatalab M Vilo J Abeygunawardena N Holloway E Kapushesky M Kemmeren P Lara GG Oezcimen A Rocca-Serra P Sansone SA ArrayExpress--a public repository for microarray gene expression data at the EBI Nucleic Acids Res 2003 31 68 71 12519949 10.1093/nar/gkg091
Fauman EB Saper MA Structure and function of the protein tyrosine phosphatases Trends Biochem Sci 1996 21 413 417 8987394 10.1016/S0968-0004(96)10059-1
Park HD Beeser AE Clancy MJ Cooper TG The S. cerevisiae nitrogen starvation-induced Yvh1p and Ptp2p phosphatases play a role in control of sporulation Yeast 1996 12 1135 1151 8896280 10.1002/(SICI)1097-0061(19960915)12:11<1135::AID-YEA11>3.0.CO;2-L
Guan K Hakes DJ Wang Y Park HD Cooper TG Dixon JE A yeast protein phosphatase related to the vaccinia virus VH1 phosphatase is induced by nitrogen starvation Proc Natl Acad Sci U S A 1992 89 12175 12179 1334559
Hanaoka N Umeyama T Ueno K Ueda K Beppu T Fugo H Uehara Y Niimi M A putative dual-specific protein phosphatase encoded by YVH1 controls growth, filamentation and virulence in Candida albicans Microbiology 2005 151 2223 2232 16000712 10.1099/mic.0.27999-0
Muda M Manning ER Orth K Dixon JE Identification of the human YVH1 protein-tyrosine phosphatase orthologue reveals a novel zinc binding domain essential for in vivo function J Biol Chem 1999 274 23991 23995 10446167 10.1074/jbc.274.34.23991
Sakumoto N Yamashita H Mukai Y Kaneko Y Harashima S Dual-specificity protein phosphatase Yvh1p, which is required for vegetative growth and sporulation, interacts with yeast pescadillo homolog in Saccharomyces cerevisiae Biochem Biophys Res Commun 2001 289 608 615 11716519 10.1006/bbrc.2001.6021
Kumar R Musiyenko A Cioffi E Oldenburg A Adams B Bitko V Krishna SS Barik S A zinc-binding dual-specificity YVH1 phosphatase in the malaria parasite, Plasmodium falciparum, and its interaction with the nuclear protein, pescadillo Mol Biochem Parasitol 2004 133 297 310 14698441 10.1016/j.molbiopara.2003.11.005
Kinoshita Y Jarell AD Flaman JM Foltz G Schuster J Sopher BL Irvin DK Kanning K Kornblum HI Nelson PS Hieter P Morrison RS Pescadillo, a novel cell cycle regulatory protein abnormally expressed in malignant cells J Biol Chem 2001 276 6656 6665 11071894 10.1074/jbc.M008536200
Lerch-Gaggl A Haque J Li J Ning G Traktman P Duncan SA Pescadillo is essential for nucleolar assembly, ribosome biogenesis, and mammalian cell proliferation J Biol Chem 2002 277 45347 45355 12237316 10.1074/jbc.M208338200
Hai T Hartman MG The molecular biology and nomenclature of the activating transcription factor/cAMP responsive element binding family of transcription factors: activating transcription factor proteins and homeostasis Gene 2001 273 1 11 11483355 10.1016/S0378-1119(01)00551-0
Wang XZ Lawson B Brewer JW Zinszner H Sanjay A Mi LJ Boorstein R Kreibich G Hendershot LM Ron D Signals from the stressed endoplasmic reticulum induce C/EBP-homologous protein (CHOP/GADD153) Mol Cell Biol 1996 16 4273 4280 8754828
Aman P Ron D Mandahl N Fioretos T Heim S Arheden K Willen H Rydholm A Mitelman F Rearrangement of the transcription factor gene CHOP in myxoid liposarcomas with t(12;16)(q13;p11) Genes Chromosomes Cancer 1992 5 278 285 1283316
Forus A Florenes VA Maelandsmo GM Fodstad O Myklebost O The protooncogene CHOP/GADD153, involved in growth arrest and DNA damage response, is amplified in a subset of human sarcomas Cancer Genet Cytogenet 1994 78 165 171 7828148 10.1016/0165-4608(94)90085-X
Yoshida H Haze K Yanagi H Yura T Mori K Identification of the cis-acting endoplasmic reticulum stress response element responsible for transcriptional induction of mammalian glucose-regulated proteins. Involvement of basic leucine zipper transcription factors J Biol Chem 1998 273 33741 33749 9837962 10.1074/jbc.273.50.33741
Haze K Yoshida H Yanagi H Yura T Mori K Mammalian transcription factor ATF6 is synthesized as a transmembrane protein and activated by proteolysis in response to endoplasmic reticulum stress Mol Biol Cell 1999 10 3787 3799 10564271
Little E Ramakrishnan M Roy B Gazit G Lee AS The glucose-regulated proteins (GRP78 and GRP94): functions, gene regulation, and applications Crit Rev Eukaryot Gene Expr 1994 4 1 18 7987045
Lee AS The glucose-regulated proteins: stress induction and clinical applications Trends Biochem Sci 2001 26 504 510 11504627 10.1016/S0968-0004(01)01908-9
Osoegawa K Mammoser AG Wu C Frengen E Zeng C Catanese JJ de Jong PJ A bacterial artificial chromosome library for sequencing the complete human genome Genome Res 2001 11 483 496 11230172 10.1101/gr.169601
Fiegler H Carr P Douglas EJ Burford DC Hunt S Scott CE Smith J Vetrie D Gorman P Tomlinson IP Carter NP DNA microarrays for comparative genomic hybridization based on DOP-PCR amplification of BAC and PAC clones Genes Chromosomes Cancer 2003 36 361 374 12619160 10.1002/gcc.10155
Forus A Florenes VA Maelandsmo GM Meltzer PS Fodstad O Myklebost O Mapping of amplification units in the q13-14 region of chromosome 12 in human sarcomas: some amplica do not include MDM2 Cell Growth Differ 1993 4 1065 1070 8117620
Wang J Meza-Zepeda LA Kresse SH Myklebost O M-CGH: Analysing microarray-based CGH experiments BMC Bioinformatics 2004 5 74 15189572 10.1186/1471-2105-5-74
Lennon G Auffray C Polymeropoulos M Soares MB The I.M.A.G.E. Consortium: an integrated molecular analysis of genomes and their expression Genomics 1996 33 151 152 8617505 10.1006/geno.1996.0177
Rogne S Myklebost O Hoyheim B Olaisen B Gedde-Dahl TJ The genes for apolipoprotein all (APOA2) and the Duffy blood group (FY) are linked on chromosome 1 in man Genomics 1989 4 169 173 2500391 10.1016/0888-7543(89)90296-6
Huang LS Bock SC Feinstein SI Breslow JL Human apolipoprotein B cDNA clone isolation and demonstration that liver apolipoprotein B mRNA is 22 kilobases in length Proc Natl Acad Sci U S A 1985 82 6825 6829 2995989
Simons A Schepens M Jeuken J Sprenger S van de Zande G Bjerkehagen B Forus A Weibolt V Molenaar I van den Berg E Myklebost O Bridge J van Kessel AG Suijkerbuijk R Frequent loss of 9p21 (p16(INK4A)) and other genomic imbalances in human malignant fibrous histiocytoma Cancer Genet Cytogenet 2000 118 89 98 10748288 10.1016/S0165-4608(99)00178-8
Parente F Grosgeorge J Coindre JM Terrier P Vilain O Turc-Carel C Comparative genomic hybridization reveals novel chromosome deletions in 90 primary soft tissue tumors Cancer Genet Cytogenet 1999 115 89 95 10598139 10.1016/S0165-4608(99)00082-5
|
16274472
|
PMC1308856
|
CC BY
|
2021-01-04 16:36:34
|
no
|
Mol Cancer. 2005 Nov 7; 4:39
|
utf-8
|
Mol Cancer
| 2,005 |
10.1186/1476-4598-4-39
|
oa_comm
|
==== Front
Mol PainMolecular Pain1744-8069BioMed Central London 1744-8069-1-351630574910.1186/1744-8069-1-35ResearchASIC3, an acid-sensing ion channel, is expressed in metaboreceptive sensory neurons Molliver Derek C [email protected] David C [email protected] Leonardo [email protected]é Michel [email protected] Frank L [email protected] Edwin W [email protected] Vollum Institute, Oregon Health and Science University, Portland, Oregon 97239, USA2 Center for Neuropharmacology and Neuroscience, Albany Medical College, Albany, NY 12208, USA3 UPMC, Dept. Medicine, Univ of Pittsburgh, Pittsburgh PA 15261, USA4 Amgen, Inc., Dept of Neuroscience, One Amgen Way. Thousand Oaks CA. 91320, USA5 Dept de Ciencias Fisiologias, Universidad del Valle, Cali, Colombia6 AstraZeneca R&D Montreal, Montreal QC H4S 1Z9, Canada2005 23 11 2005 1 35 35 9 9 2005 23 11 2005 Copyright © 2005 Molliver et al; licensee BioMed Central Ltd.2005Molliver et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
ASIC3, the most sensitive of the acid-sensing ion channels, depolarizes certain rat sensory neurons when lactic acid appears in the extracellular medium. Two functions have been proposed for it: 1) ASIC3 might trigger ischemic pain in heart and muscle; 2) it might contribute to some forms of touch mechanosensation. Here, we used immunocytochemistry, retrograde labelling, and electrophysiology to ask whether the distribution of ASIC3 in rat sensory neurons is consistent with either of these hypotheses.
Results
Less than half (40%) of dorsal root ganglion sensory neurons react with anti-ASIC3, and the population is heterogeneous. They vary widely in cell diameter and express different growth factor receptors: 68% express TrkA, the receptor for nerve growth factor, and 25% express TrkC, the NT3 growth factor receptor. Consistent with a role in muscle nociception, small (<25 μm) sensory neurons that innervate muscle are more likely to express ASIC3 than those that innervate skin (51% of small muscle afferents vs. 28% of small skin afferents). Over 80% of ASIC3+ muscle afferents co-express CGRP (a vasodilatory peptide). Remarkably few (9%) ASIC3+ cells express P2X3 receptors (an ATP-gated ion channel), whereas 31% express TRPV1 (the noxious heat and capsaicin-activated ion channel also known as VR1). ASIC3+/CGRP+ sensory nerve endings were observed on muscle arterioles, the blood vessels that control vascular resistance; like the cell bodies, the endings are P2X3- and can be TRPV1+. The TrkC+/ASIC3+ cell bodies are uniformly large, possibly consistent with non-nociceptive mechanosensation. They are not proprioceptors because they fail two other tests: ASIC3+ cells do not express parvalbumin and they are absent from the mesencephalic trigeminal nucleus.
Conclusion
Our data indicates that: 1) ASIC3 is expressed in a restricted population of nociceptors and probably in some non-nociceptors; 2) co-expression of ASIC3 and CGRP, and the absence of P2X3, are distinguishing properties of a class of sensory neurons, some of which innervate blood vessels. We suggest that these latter afferents may be muscle metaboreceptors, neurons that sense the metabolic state of muscle and can trigger pain when there is insufficient oxygen.
==== Body
Background
ASICs are ion channels that open when the extracellular pH drops [1,2]. They belong to the epithelial sodium channel/degenerin superfamily of ion channels [3,4]. Members of this family are selective for sodium, are inhibited by the diuretic drug amiloride, and have two membrane-spanning regions with intracellular N- and C-termini and a large extracellular loop; multiple subunits must assemble to form a functional ion channel. There are three known ASIC genes that can form ion channels and two splice variants. The mRNA for all these ASICs are found in sensory ganglia; ASIC3 is unique in that, in rodents, little or none of it is expressed outside of sensory neurons [5].
The present study asks whether the distribution of ASIC3 among subsets of sensory neurons is consistent with its hypothesized roles as a transducer of two distinctly different sensations: ischemic pain and mechanosensation. ASICs were first considered as possible mechanosensitive ion channels because of their structural similarity to degenerins, which subserve touch sensitivity in Caenorhabditis elegans [6]. In rodents, ASICs are expressed in specific skin mechanosensory endings [7-9] and ASIC knockout mice exhibit subtle behavioral changes in their touch sensitivity in some assays [9-11], but not in others [12,13]. ASIC3 may act as a sensor of the lactic acidosis that occurs during ischemic pain because: a) it generates exceedingly large currents in dorsal root ganglion (DRG) sensory neurons that innervate the heart [14,15], an organ whose only conscious sensation is ischemic pain (angina); b) it responds to the small change in extracellular pH (a drop from 7.4 to 7.0) that occurs in the extracellular space of metabolically stressed muscle [15]; c) it responds better to lactic acid than to other forms of acid [16]; d) multiple injections of acid into skeletal muscle induce a long-lived painful hypersensitivity, but this does not occur in mice that lack ASIC3 [17]. In contrast to a possible role in muscle pain, evidence from transgenic mice argues against ASIC3 directly triggering cutaneous pain [18], although ASICs appear to contribute to acid-induced pain in human skin [19,20].
Two prior studies explored ASIC3 immunoreactivity in sensory neurons. One found ASIC3 and ASIC2 to co-localize in dorsal root ganglion neurons and to be present more often in larger neurons than in smaller [21]. The other found ASIC3 in both large and small neurons of the trigeminal ganglion and to exhibit some co-expression with CGRP [22]. The present study extends these earlier ones by combining retrograde labeling, electrophysiology and immunocytochemistry to distinguish various populations of nociceptive and non-nociceptive sensory neurons. The results suggest that high ASIC3 expression is an essential feature of a specific subset of nociceptors that respond to the metabolic state of muscle, but this is not the only population of sensory neurons that express ASIC3.
Results
ASIC3 observed in both TrkA and TrkC-positive neurons
The guinea pig ASIC3 antibody from Neuromics was used for all staining of dorsal root ganglia (DRG, Figs. 1, 2, 3, 4, 5, 6). We tested its specificity in COS-7 cells transfected with cDNA for the following ASICs: 1a, 1b, 2a, 3. It labelled ASIC3-transfected cells but not the others (Fig. 1A shows ASIC3 and ASIC1a; others are not shown but were negative like ASIC1a). No staining occurred if the antibody was preincubated with antigenic peptide or if the secondary antibody was applied without the primary (not shown). The antibody appears free of non-specific staining at the dilutions we used because it does not label tissue in the brainstem (Fig. 3B), which has no reported ASIC3 mRNA. Both guinea pig and rabbit anti-ASIC3 were used to label vascular innervation (Figure 7) and gave indistinguishable results.
Figure 1 Immunocytochemical localization of ASIC3 in rat sensory neurons. A, ASIC3 antibody labeled COS-7 cells transfected with ASIC3 (top), but not those transfected with ASIC1a (bottom). It also failed to label cells transfected with ASIC1b and ASIC2a (not shown); ASIC2b and ASIC4 were not tested. B, ASIC3 immunoreactivity in a section of L5 DRG. Less than half of neurons are labeled and no non-neurons are labeled. Note the presence of both small and large labeled neurons (inset). Scale bar, 40 μm. C, Size-frequency distribution of ASIC3-positive neurons (black) and of all neurons (grey) from the L5 DRG. ASIC3+ cells are present in all size bins. 3345 neurons (12–15 sections from 3 rats) were counted, of which 1359 (40%) were ASIC3-positive.
Figure 2 ASIC3 co-expression with TrkA and TrkC. Photomicrographs of DRG neurons labeled with antibodies to ASIC3 (red) and TrkA (A, green) or TrkC (B, green). Double-labeled cells appear yellow or orange. Circle charts show the fraction of double labeled cells (yellow), unlabeled (black), ASIC3-only (red), and TrkA or TrkC-only (green). 944 cells counted for A; 832 for B. Scale bar, 40 μm.
Figure 3 ASIC3 is not expressed by muscle spindle afferents. A, Section of DRG labeled for ASIC3 (red) and for the calcium-binding protein parvalbumin (green), which labels muscle spindle afferents. Only 2 of 1400 parvalbumin-positive cells had detectable ASIC3 staining. B, Section of brainstem containing a string of sensory neurons in the mesencephalic nucleus of the 5th nerve (MesV) stained with either an antibody to a neurofilament protein (N52) to reveal all the sensory neurons (left, string of large cells arranged vertically; top punctuate labeling are MesV axons in cross section) or with the ASIC3 antibody (right). No MesV cells expressed ASIC3 (n = 3 rats). "IV" marks the fourth ventricle. Scale bars, 40 μm.
Figure 4 ASIC3 expression compared to other nociceptive ion channels. DRG sections double-labeled for ASIC3 (red) and either P2X3 (green, A) or TRPV1 (green, B); double-labeled cells are yellow or orange. Circle charts show the relative distribution of the three receptors among all DRG neurons. ASIC3 and P2X3 expression overlapped in only 4% of neurons, while some 35% of ASIC3-positive neurons also express TRPV1. The rare cells positive for both ASIC3 and P2X3 were large (eg. cell at lower left in A). 1406 cells counted for A; 995 for B. C, D, electrophysiological recordings reveal a qualitatively similar expression pattern for currents conforming to ASIC3, P2X3, or TRPV1. The cell that generated the currents in C counts as a co-expressor of ASIC3 and P2X3, fitting into the yellow bin in the circle plot. The cell in D is a co-expressor of ASIC3 and TRPV1. Cells were counted positive if they had at least 0.3 nA of current in response to pH 6.8 (ASIC3+), 50 μM ATP (P2X3+; transient current only), or 1 μM capsaicin (TRPV1). 182 cells recorded for C; 169 for D.
Figure 5 Muscle nociceptors more likely express ASIC3 than do skin nociceptors. A, Photomicrographs on the left and right are the same DRG section photographed for ASIC3 immunoreactivity (right) and fluorogold (left) that was retrogradely transported from injections into muscle (top) or skin (bottom). Arrows indicate cells that are 25 μm or smaller and contain both fluorogold and ASIC3 immunoreactivity. Scale bar, 40 μm. B, The percent of small (<25 μm) muscle or skin afferents that stain for ASIC3 (Immuno) or exhibit more than 0.3 nA of current evoked by a step to pH 6.8 from pH 7.4 (Ephys). 37 and 32 small neurons were recorded for skin and muscle, respectively. 117 and 113 small neurons were counted for skin and muscle immunocytochemistry.
Figure 6 Most muscle afferents that express ASIC3 also express CGRP. Three photomicrographs of the same DRG section showing: fluorogold transported from a muscle injection (top); ASIC3 immunoreactivity (middle); CGRP immunoreactivity (bottom). Double arrowheads in the bottom image indicate cells positive for each label. Arrows indicate cells positive for fluorogold and for ASIC3, but not for CGRP. 83% (168/202) of ASIC3-positive muscle afferents were CGRP-positive (373 total muscle afferents counted). Scale bar, 40 μm.
Figure 7 Double-label immunofluorescence of arterial innervation among muscles in the lateral plantar compartment of the rat hind paw. Red labeling is with secondary antibodies conjugated with Cy3, green with Cy2. The images in A and B are from a larger source artery and those in C-H are from smaller tributaries. L = arterial lumen, TA = tunica adventitia, TM = tunica media, M = skeletal muscle. Scale bar = 25 μm. A, B. Cross section of a larger artery with lumen (L) toward the right. Anti-CGRP (A) labeled sensory innervation that is located primarily in the tunica adventitia (green arrows). This CGRP-positive sensory innervation is supplied by axons that have thin calibers (small green arrows) and slightly thicker-calibers (large green arrows). Double labeling with anti-NPY (B) revealed sympathetic innervation concentrated at the border between the tunica media and tunica adventitia (red arrows). C, D. Cross section of an arteriole with lumen to the upper right, Muscle fibers (M) in cross section below. Double staining with anti-ASIC3 (C) and anti-NF200 (D) demonstrated that ASIC3 is expressed primarily on the slightly thicker caliber innervation that is NF200-positive (yellow arrows). E, F. Small arterioles viewed both in cross section (lumen labelled L) and on edge (oval structure labelled TM). Double labeling with anti-ASIC3 (E) and anti-CGRP (F) revealed that ASIC3 is coexpressed exclusively with CGRP, primarily on the thicker-caliber innervation (yellow arrows). The smaller-caliber CGRP innervation is usually ASIC3-negative (green arrows). G, H. Edge of an artery cross section; lumen is out of the picture to the right; tunica media (TM) and tunica adventitia (TA) are marked. Double labeling with anti-TRPV1 (E) and anti-CGRP (F) demonstrated that TRPV1 is also coexpressed on the larger caliber CGRP-positive innervation (yellow arrows).
Some 40% (1359/3345 cells in single label studies; 1932/4783 in double label) of DRG neurons stain for ASIC3 and there is no evident staining in non-neuronal cells or axon tracts (Fig. 1B). Both small (<25 μm diameter) and large (>40 μm) diameter sensory neurons express ASIC3. This contrasts with TRPV1, which is almost exclusively in smaller neurons [23]. The wide size distribution (Fig. 1C) suggests that anti-ASIC3 may label neurons that differ in conduction velocity and perhaps in sensory modality [24-26].
The TrkA receptor (the binding site for nerve growth factor) and the TrkC receptor (binding site for NT3 growth factor) label non-overlapping populations of sensory neurons [27,28]. The majority (68%, 284/417) of ASIC3+ neurons co-expressed TrkA (Fig. 2A). TrkA+ sensory neurons in adult rodents are considered to be NGF-sensitive nociceptors [29-31].
A minority (25%, 93/373) of ASIC3+ neurons co-expressed TrkC (Fig. 2B), a growth factor receptor used by many non-nociceptive mechanosensors [30]. Proprioceptors are among the sensory neurons that express TrkC [32]. Proprioceptors have two other testable properties: they express parvalbumin [33] and those from masticatory muscles have cell bodies in the brainstem at the mesencephalic trigeminal nucleus [34]. Only 2 of 1400 parvalbumin+ neurons co-expressed ASIC3 (Fig. 3A) and there was no staining in the mesencephalic nucleus (Fig. 3B). Evidently, ASIC3+/TrkC+ cells are not proprioceptors.
ASIC3 co-expression with two nociceptive ion channels
To characterize the nociceptive population of neurons containing ASIC3, we tested for co-expression with two other ion channels thought to be pain-transducers: P2X3, an ATP-gated ion channel [35], and TRPV1, the capsaicin receptor that responds to noxious heat [36]. There was little overlap of ASIC3 and P2X3: although many neurons stained for one or the other, less than 5% (51/1406) stained strongly for both (Fig. 4A). We used electrophysiological recordings to test for expression of functional protein; only 10% of cells (18/182) had both ASIC3-like and P2X3-like currents that were larger than 0.3 nA (Fig. 4C). The cells that co-expressed ASIC3 and P2X3 were all large (examples: the one orange cell in Fig. 4A, lower left, and the cell (48 μm diameter) that gave the traces in Fig. 4C). Interestingly, virtually every large P2X3+ cell also stained positive for ASIC3, suggesting that this may be a distinct, albeit small population of sensory neurons.
There is more overlap between ASIC3 and TRPV1. 12% (122/995) of all cells stained for both, and 21% (35/169) had substantial current carried by both; 31% (122/388) of ASIC3+ cells were TRPV1+ (Fig. 4B,D). ASIC3/TRPV1 co-expressors were small (yellow cells, Fig. 4B), unlike the ASIC3/P2X3 co-expressors.
More ASIC3 expression in small muscle afferents than small skin afferents
Although ischemic pain occurs when there is low blood flow to muscle, low flow to skin occurs routinely without any pain — for example, for thermoregulation or during the fight or flight response. Therefore, one might predict that, if ASIC3 is the sensor for ischemic pain, it should be in more muscle nociceptors than skin nociceptors. We tested this prediction by labelling muscle or skin afferents with retrogradely transported dyes, and then studying them with either electrical recordings or ASIC3 antibody staining; diI was used for electrophysiology and fluorogold for immunocytochemistry. In order to focus on nociception, we analyzed data for cells less than 25 μm diameter. However, as noted before, ASIC3 is also seen in many large neurons.
Figure 5 shows that small muscle afferents are more likely to express ASIC3 than are small skin afferents. A stimulus of pH 6.8, chosen to selectively stimulate ASIC3, elicited large current (>0.3 nA) in 56% (18/32) of small muscle afferents, but in only 11% (4/37) of small skin afferents (Fig. 5B). Immunocytochemistry presents a similar picture. The arrows in the right images of Fig. 5A point to small neurons (<25 μm) positive for both ASIC3 (right) and for fluorogold (left) transported from muscle (upper) or skin (lower); 50% (57/113) of small muscle afferents stained for ASIC3, compared to 28% (33/117) of small skin afferents (Fig. 5B).
ASIC3 and CGRP co-express on sensory endings innervating blood vessels
Whether they have small diameters or large, most muscle afferents (83%, 168/173) that express ASIC3 also express the vasodilatory peptide, CGRP. The double arrowheads in the bottom image of Fig. 6 point to cells with evident nuclei that stained for all three labels: a) fluorogold that was previously injected into muscle (top panel); b) ASIC3 (middle); and c) CGRP (bottom). The arrows point to three ASIC3-positive muscle afferents that did not stain for CGRP. Overlap of CGRP and ASIC3 was also evident in skin afferents: 69% (168/244) of ASIC3 skin afferents also stained with the anti-CGRP.
Co-expression of ASIC3 with a vasodilatory peptide suggests that they might be expressed on sensory endings that innervate vasculature. Fig. 7 shows immunofluorescence images of arteries within rat lateral plantar muscles using the rabbit anti-ASIC3. L indicates the artery lumen, TA the tunica adventitia, and TM the tunica media. Nerve fibers that express CGRP are from sensory neurons [37]; they are relatively sparse and are present in the outer, adventitial layer (TA) of the artery wall (Fig. 7A). Fibers containing the neuropeptide NPY are from sympathetic neurons [37]; these fibers are denser than the CGRP fibers and lie at the border between the tunica adventitia and tunica media (TM) parts of the wall (Fig. 7B). Although all of the arterial innervation has a relatively small caliber, some CGRP-positive axons are noticeably thicker and are labeled by anti-200 kD neurofilament protein (NF200) and anti-myelin basic protein; this indicates that they are myelinated Aδ fibers [37]. The thinner CGRP-positive arterial innervation lacked NF200 and myelin basic protein immunoreactivity indicating that they are unmyelinated C fibers [37].
The ASIC3 antibody labels the thicker CGRP-expressing fibers in the tunica adventitia (large arrows, Fig. 7E,F) and it is co-expressed with NF200 (Fig. 7C,D). The thinner CGRP-expressing axons have little or no detectable ASIC3 immunoreactivity (small arrows, Fig. 7E,F). TRPV1 is also expressed on at least some of the thicker caliber CGRP-positive axons (Fig. 7G,H). In contrast to TRPV1 and ASIC3, antibodies to P2X3 do not label vascular afferents (not shown). In the adult, vascular CGRP-positive axons express TrkA but not TrkC [28].
Discussion
We find ASIC3 in less than half of all DRG sensory neurons, but the population is diverse: ASIC3 can be in both small neurons and large, TrkA-expressors and TrkC. The neurons fall into at least two distinct subsets: 1) small to medium-size, TrkA- and CGRP-positive muscle afferents that lack P2X3 and can innervate small arteries; and 2) large neurons that lack parvalbumin and can express either TrkA or TrkC. We discuss the possibilities that neurons in the first subset are muscle metaboreceptors and the second includes non-proprioceptive mechanosensors.
Metaboreception and the axon reflex
Because ASIC3 can open with quite small (0.5 unit) pH changes [15] and responds better to lactic acid than to other acids [16], it seems suitable for detecting when muscle is using anaerobic metabolism. Some consider metaboreception to be a subset of nociception [38]. In this view, low levels of metaboreceptor activity trigger compensatory responses in peripheral tissues and high levels trigger pain [39-41]. If ASIC3 is a molecular sensor for ischemia, it should be present in all cardiac DRG afferents (which, though diverse, all respond to cardiac ischemia, [42]), and it should be relatively enriched in sensory neurons that innervate muscle compared to skin because ischemic pain is characteristic of muscle but not skin. Our data here (on skin and muscle) and previously (on cardiac afferents, [14]) fulfill these expectations.
We found profound overlap of ASIC3 and the vasodilatory peptide CGRP. Over 80% of ASIC3-positive neurons that innervate muscle also express CGRP. Importantly, many sensory axons that innervate small arteries express ASIC3 together with CGRP and TRPV1. The axons that co-express the three markers appear to be lightly myelinated Aδ fibers because they express NF200, a marker of myelinated axons [43], whereas we saw no evidence that CGRP-positive small-caliber (C) arterial axons express ASIC3 or TRPV1. These ASIC3/TRPV1/CGRP-positive cells and axons uniformly fail to express P2X3, the primary ATP-gated ion channel in sensory neurons. Finally, the phenotype seems not to be unique to rats because ASIC3 labeling also occurs on CGRP-positive Aδ fibers that terminate on small arteries and arterioles in monkeys (Paré and Rice, unpublished). Extensive overlap of ASIC3 and CGRP has previously been reported in trigeminal ganglion neurons [22].
Release of CGRP in the periphery clearly contributes to the "axon reflex", in which activation of C fibers causes vasodilation and extravasation [44,45]. The tight co-expression of ASIC3, TRPV1 and CGRP on muscle arterial afferents raises the possibility that they function together as sensors and effector, increasing local blood flow in response to the presence of lactic acid and elevated muscle temperature. Co-expression of ASIC3 and TRPV1 should provide a cell with a broad range of acid sensitivity.
These observations add to a growing literature describing the phenotype of arterial afferents. They express TrkA immunoreactivity but not TrkC in adult mice and rats, though they express Trk C transiently during development [28]. The larger caliber arterial afferents express the histamine receptor 3, which is implicated in intense pressure mechanical nociception and allodynia [46].
We did not test for co-expression of ASIC3 with other ASIC subunits, but the existing literature addresses this well. ASIC3 and ASIC2, particularly the 2b splice variant, exhibit strong overlap [21]. ASIC1 appears expressed in a broader population of sensory neurons than is ASIC3 [21] and it is observed in many CGRP-expressing neurons [47]. The implication that ASIC3 and ASIC1 are often in the same cells is confirmed by both electrical and immunocytochemical observations in transgenic mice [48].
Mechanosensation
Two labs have independently shown that ASICs are present in specific non-nociceptive mechanosensory nerve endings [7,9]. This includes ASIC3, with immunoreactivity observed in Aβ innervation that terminate as Merkel endings, Meissner corpuscle endings and hair follicle lanceolate endings in mice [10], as well as in rats and monkeys (Paré and Rice, unpublished). Data here indicates that 25% of ASIC3-staining cells also express TrkC, a growth factor receptor generally associated with non-nociceptive mechanosensors [30] and most clearly proven present in muscle proprioceptors [32]. However, the full range of sensory modalities among TrkC expressors is unknown. While it seems likely that the ASIC3+/TrkC+ cells include non-nociceptive mechanosensors, we cannot be certain.
We do know for certain that the ASIC3+/TrkC+ cells are not muscle spindle proprioceptors because they are absent from the mesencephalic trigeminal nucleus and they do not express parvalbumin. Other non-nociceptive mechanosensors show a diversity of expression patterns. Merkel cells and lanceolate endings do not express parvalbumin [49,50], lanceolate endings express CGRP [28,49-51], and Meissner's corpuscles express both CGRP and P2X3 [52]. Each of these expression patterns is observed in at least some of the ASIC3+ cells that we report. Thus, we have no data that conflicts with the expression pattern reported for skin mechansensory endings by Price and colleagues [10].
Conclusion
Although less than half of all DRG sensory neurons express ASIC3, it is in multiple populations. One population has properties expected of metaboreceptors — cells that detect the metabolic state of muscle, mediate compensatory reflexes, and may initiate ischemic muscle pain. Over 80% of the ASIC3-containing muscle afferents co-express the vasodilatory peptide CGRP and some of these innervate small arteries; these cells seem poised to modulate blood flow in response to muscle stress. Another population of ASIC3-containing cells are quite large, can express TrkC, but do not express parvalbumin; this is a pattern observed with some non-nociceptive mechanoreceptors, but may not be unique to them.
Methods
Immunocytochemistry
Young adult Sprague-Dawley rats (150 g) were deeply anesthetized and perfused transcardially with 0.15 M phosphate-buffered saline, pH7.4 (PBS: 150 mM NaCl, 25 mM Na2HPO4, pH 7.4) followed by approximately 300 ml 3% paraformaldehyde with 15% picric acid in PBS or by 4% paraformaldehyde. The quality of ASIC3 staining was reduced with stronger fixation and in older rats, which exhibited unacceptable background fluorescence. Tissues were immediately dissected and placed in 30% sucrose in PBS overnight, then rapidly frozen and stored at -80°C until needed. Fourteen μm sections were cut on a cryostat and mounted on glass slides. Slides were placed in blocking solution containing 2% normal donkey serum, 0.2% Triton X-100 in PBS for 30 minutes, then incubated in primary antibody solution overnight at 4°C. Table 1 gives dilutions of the various antisera that were used. Next, slides were washed 3 times for 3 minutes in PBS and incubated in secondary antibodies diluted 1:200 in blocking solution for 30 minutes. Secondary antibodies were donkey anti-rabbit, donkey anti-mouse and donkey anti-guinea pig conjugated to CY3 or CY2 (Jackson Immunoresearch, West Grove PA). Slides were washed 3 times in PBS, mounted on coverslips in PBS and immediately photographed. For quantification, sections were taken from the fifth lumbar (L5) DRG. L5 was chosen because it is the ganglion used most commonly for immunocytochemistry. To minimize over-counting of large cells, only cells showing a clear nucleus were counted. All histograms and circle charts involved counts of over 1000 cells from at least three rats. Immunofluorescence assessments of arterial innervation were conducted on the lateral plantar artery and arterioles in the lateral plantar compartment of the foot. As described previously in detail [52], double labeling was conducted with various combinations of primary antibodies raised in different species.
Table 1 Antisera.
Antisera Species Dilution Source
ASIC3 guinea pig
rabbit 1:300
1:1000 Neuromics
Robert Elde, University of Minnesota, Minneapolis
P2X3 rabbit 1:2000 Neuromics
TRPV1 rabbit
guinea pig 1:1000
1:1000 Neuromics
Neuromics
TrkA rabbit 1:300 Chemicon, Temecula, CA
TrkC Goat 1:300 Louis Reichardt, Universityof California, SF
Parvalbumin mouse 1:1000 Sigma, St. Louis, MO
N52 mouse 1:10000 Sigma
CGRP mouse
rabbit
guinea pig 1:2000
1:1000
1:500 RBI, Natick, MA
Chemicon
Peninsula
NPY Sheep 1:1000 Chemicon
NF200 rabbit 1:1000 Chemicon
The guinea pig anti-ASIC3 was used to label all DRG (Figs. 1-6). Both guinea pig and rabbit anti-ASIC3 were used to label vascular innervation; when used together, each blocked labelling by the other. Fig. 7 shows results with the rabbit anti-ASIC3.
The Neuromics guinea pig ASIC3 antibody was used for all DRG cell body experiments (Figs. 1, 2, 3, 4, 5, 6). Both the guinea pig antibody and the rabbit ASIC3 antibody (generous gift of Robert Elde lab) were used to label vascular innervation (Figure 7 shows results with the rabbit antibody). The presence of either antibody blocked labelling by the other (not shown). The antibodies were raised against extracellular residues (237–257 for rabbit; 285–304 for guinea pig).
Retrograde Labeling
DRG neurons innervating muscle and skin were labeled by injecting fluorescent dyes into either quadriceps muscle [53] or intradermally in the upper back between the shoulder blades. The intradermal injection raises a small blister that disappears within an hour; we did this in back skin because rats scratch and bite blisters put elsewhere, which would confound the assumption that dye is confined to the injection site. DiI (Molecular Probes, Eugene, OR. 5% in DMSO) was used for electrophysiology [54] and fluorogold (Fluorochrome Inc, Englewood CO. 1% in water) for immunocytochemistry because it withstands membrane permeabilization with detergent [55]. Dye was injected one to two weeks prior to sacrificing the rat.
Electrophysiology
Whole cell patch clamp of dissociated sensory neurons measured currents carried by ASIC3, P2X3, or vanilloid receptors at -70 mV holding potential, room temperature. Solutions perfused the cells through 10 μl capillary pipettes and were changed within 20 msec with a computer-driven solenoid valve system. We used concentrations of ATP (50 μM) and capsaicin (1 μM) sufficient to evoke maximal responses in P2X3 and vanilloid receptors, but used a submaximal pH (6.8) stimulus. Because ASIC3 is more sensitive than other rat ASICs [15], pH 6.8 triggers ASIC3 channels without substantial activation of other subtypes. P2X3 currents were distinguished from other P2X receptors through their rapid desensitization. ATP-Na2 was added to the external solution from a 50 mM stock prepared in water at pH 7.4 (stored at -20°C). Capsaicin was prepared daily in ethanol.
Cervical or lumbar DRG were dissected and dissociated as previously described [14], plated on glass coverslips coated with poly-d-lysine and laminin, and maintained for approximately 2 hr at 37°C at 5% CO2/95% air in F12 media (Gibco BRL) plus 50 ng/ml nerve growth factor (NGF, Biomedical Tech. Inc.). Media was then changed to L15 (Gibco BRL) plus 50 ng/ml NGF at 23°C in humidified air. Electrophysiological recordings were made 24–48 hours after dissociation. Patch pipettes were pulled from borosilicate glass (Garner Glass, Claremont, CA) to 2.5–4 MΩ resistance. The standard internal solution contained (in mM): 135 methane sulphonic acid, 150 KOH, 10 KCl, 8 NaCl, 1 MgCl2, 10 MOPS, 0.3 Na3-GTP, 2 ATP-Mg, and 0.5 EGTA, adjusted to pH 7.0 with KOH. The control external solution contained (in mM): 140 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2 10 HEPES, 10 MES, and 5 glucose, adjusted to pH 7.4 with n-methyl-d-glucamine; this was adjusted to pH 6.8 with HCl for the ASIC3-evoking solution. Data acquisition and analysis used PClamp 6.02 (Axon Instruments, Inc.) and Origin 6 (Microcal Software Inc). Chemicals were from Sigma Co. unless otherwise stated.
Abbreviations
ASIC, acid sensing ion channel; ASIC3ir, ASIC3 immunoreactivity; CGRP, neuropeptide in sensory neurons that dilates blood vessels; DiI, a red lipophilic dye used for retrograde labeling of neurons; DRG, dorsal root ganglion; N200, neurofilament protein present mostly in myelinated axons; N52, neurofilament protein present in all sensory axons; NGF, nerve growth factor; NPY, neuropeptide Y; NT3, neurotrophin-3; ret, receptor for glial derived neurotrophic factor (GDNF); P2X3, ion channel opened by ATP; Trk A, nerve growth factor receptor; Trk C, neurotrophin-3 receptor; TRPV1 and VR1, ion channel opened by capsaicin, heat, or low pH; TA, tunica adventia; TM, tunica media
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
DCM and DCI performed the immunocytochemistry on dorsal root ganglia. LF performed the electrophysiology. MP and FLR performed immunocytochemistry on vasculature. EWM oversaw the research and prepared the manuscript with help from all others.
Acknowledgements
The work was supported by NIH research grants to EWM and FLR and NRSA fellowships to DCM and DCI.
==== Refs
Krishtal O The ASICs: signaling molecules? Modulators? Trends Neurosci 2003 26 477 483 12948658 10.1016/S0166-2236(03)00210-8
Waldmann R Proton-gated cation channels--neuronal acid sensors in the central and peripheral nervous system Adv Exp Med Biol 2001 502 293 304 11950145
Bianchi L Driscoll M Protons at the gate: DEG/ENaC ion channels help us feel and remember Neuron 2002 34 337 340 11988165 10.1016/S0896-6273(02)00687-6
Kellenberger S Schild L Epithelial sodium channel/degenerin family of ion channels: a variety of functions for a shared structure Physiol Rev 2002 82 735 767 12087134
Waldmann R Lazdunski M H(+)-gated cation channels: neuronal acid sensors in the NaC/DEG family of ion channels Curr Opin Neurobiol 1998 8 418 424 9687356 10.1016/S0959-4388(98)80070-6
Ernstrom GG Chalfie M Genetics of sensory mechanotransduction Annu Rev Genet 2002 36 411 453 12429699 10.1146/annurev.genet.36.061802.101708
Garcia-Anoveros J Samad TA Zuvela-Jelaska L Woolf CJ Corey DP Transport and localization of the DEG/ENaC ion channel BNaC1alpha to peripheral mechanosensory terminals of dorsal root ganglia neurons J Neurosci 2001 21 2678 2686 11306621
Duggan A Garcia-Anoveros J Corey DP The PDZ domain protein PICK1 and the sodium channel BNaC1 interact and localize at mechanosensory terminals of dorsal root ganglion neurons and dendrites of central neurons J Biol Chem 2002 277 5203 5208 11739374 10.1074/jbc.M104748200
Price MP Lewin GR McIlwrath SL Cheng C Xie J Heppenstall PA Stucky CL Mannsfeldt AG Brennan TJ Drummond HA Qiao J Benson CJ Tarr DE Hrstka RF Yang B Williamson RA Welsh MJ The mammalian sodium channel BNC1 is required for normal touch sensation Nature 2000 407 1007 1011 11069180 10.1038/35039512
Price MP McIlwrath SL Xie J Cheng C Qiao J Tarr DE Sluka KA Brennan TJ Lewin GR Welsh MJ The DRASIC cation channel contributes to the detection of cutaneous touch and acid stimuli in mice Neuron 2001 32 1071 1083 11754838 10.1016/S0896-6273(01)00547-5
Page AJ Brierley SM Martin CM Martinez-Salgado C Wemmie JA Brennan TJ Symonds E Omari T Lewin GR Welsh MJ Blackshaw LA The ion channel ASIC1 contributes to visceral but not cutaneous mechanoreceptor function Gastroenterology 2004 127 1739 1747 15578512 10.1053/j.gastro.2004.08.061
Roza C Puel JL Kress M Baron A Diochot S Lazdunski M Waldmann R Knockout of the ASIC2 channel in mice does not impair cutaneous mechanosensation, visceral mechanonociception and hearing J Physiol 2004 558 659 669 15169849 10.1113/jphysiol.2004.066001
Drew LJ Rohrer DK Price MP Blaver KE Cockayne DA Cesare P Wood JN Acid-sensing ion channels ASIC2 and ASIC3 do not contribute to mechanically activated currents in mammalian sensory neurones J Physiol 2004 556 691 710 14990679 10.1113/jphysiol.2003.058693
Benson CJ Eckert SP McCleskey EW Acid-evoked currents in cardiac sensory neurons: A possible mediator of myocardial ischemic sensation Circ Res 1999 84 921 928 10222339
Sutherland SP Benson CJ Adelman JP McCleskey EW Acid-sensing ion channel 3 matches the acid-gated current in cardiac ischemia-sensing neurons Proc Natl Acad Sci U S A 2001 98 711 716 11120882 10.1073/pnas.011404498
Immke DC McCleskey EW Lactate enhances the acid-sensing Na+ channel on ischemia-sensing neurons Nat Neurosci 2001 4 869 870 11528414 10.1038/nn0901-869
Sluka KA Price MP Breese NM Stucky CL Wemmie JA Welsh MJ Chronic hyperalgesia induced by repeated acid injections in muscle is abolished by the loss of ASIC3, but not ASIC1 Pain 2003 106 229 239 14659506 10.1016/S0304-3959(03)00269-0
Chen CC Zimmer A Sun WH Hall J Brownstein MJ Zimmer A A role for ASIC3 in the modulation of high-intensity pain stimuli Proc Natl Acad Sci U S A 2002 99 8992 8997 12060708 10.1073/pnas.132076299
Ugawa S Ueda T Ishida Y Nishigaki M Shibata Y Shimada S Amiloride-blockable acid-sensing ion channels are leading acid sensors expressed in human nociceptors J Clin Invest 2002 110 1185 1190 12393854 10.1172/JCI200215709
Jones NG Slater R Cadiou H McNaughton P McMahon SB Acid-induced pain and its modulation in humans J Neurosci 2004 24 10974 10979 15574747 10.1523/JNEUROSCI.2619-04.2004
Alvarez de la Rosa D Zhang P Shao D White F Canessa CM Functional implications of the localization and activity of acid-sensitive channels in rat peripheral nervous system Proc Natl Acad Sci U S A 2002 99 2326 2331 11842212 10.1073/pnas.042688199
Ichikawa H Sugimoto T The co-expression of ASIC3 with calcitonin gene-related peptide and parvalbumin in the rat trigeminal ganglion Brain Res 2002 943 287 291 12101052 10.1016/S0006-8993(02)02831-7
Caterina MJ Julius D The vanilloid receptor: a molecular gateway to the pain pathway Annu Rev Neurosci 2001 24 487 517 11283319 10.1146/annurev.neuro.24.1.487
Harper AA Lawson SN Conduction velocity is related to morphological cell type in rat dorsal root ganglion neurones J Physiol 1985 359 31 46 3999040
Hoheisel U Mense S Observations on the morphology of axons and somata of slowly conducting dorsal root ganglion cells in the cat Brain Res 1987 423 269 278 3676809 10.1016/0006-8993(87)90849-3
Djouhri L Lawson SN Abeta-fiber nociceptive primary afferent neurons: a review of incidence and properties in relation to other afferent A-fiber neurons in mammals Brain Res Brain Res Rev 2004 46 131 145 15464202 10.1016/j.brainresrev.2004.07.015
McMahon SB Armanini MP Ling LH Phillips HS Expression and coexpression of Trk receptors in subpopulations of adult primary sensory neurons projecting to identified peripheral targets Neuron 1994 12 1161 1171 7514427 10.1016/0896-6273(94)90323-9
Funfschilling U Ng YG Zang K Miyazaki J Reichardt LF Rice FL TrkC kinase expression in distinct subsets of cutaneous trigeminal innervation and nonneuronal cells J Comp Neurol 2004 480 392 414 15558783 10.1002/cne.20359
Smeyne RJ Klein R Schnapp A Long LK Bryant S Lewin A Lira SA Barbacid M Severe sensory and sympathetic neuropathies in mice carrying a disrupted Trk/NGF receptor gene Nature 1994 368 246 249 8145823 10.1038/368246a0
Snider WD McMahon SB Tackling pain at the source: new ideas about nociceptors Neuron 1998 20 629 632 9581756 10.1016/S0896-6273(00)81003-X
Barbacid M Neurotrophic factors and their receptors Curr Opin Cell Biol 1995 7 148 155 7612265 10.1016/0955-0674(95)80022-0
Klein R Silos-Santiago I Smeyne RJ Lira SA Brambilla R Bryant S Zhang L Snider WD Barbacid M Disruption of the neurotrophin-3 receptor gene trkC eliminates la muscle afferents and results in abnormal movements Nature 1994 368 249 251 8145824 10.1038/368249a0
Celio MR Calbindin D-28k and parvalbumin in the rat nervous system Neuroscience 1990 35 375 475 2199841 10.1016/0306-4522(90)90091-H
Cody FW Lee RW Taylor A A functional analysis of the components of the mesencephalic nucleus of the fifth nerve in the cat J Physiol 1972 226 249 261 4263681
Hamilton SG McMahon SB ATP as a peripheral mediator of pain J Auton Nerv Syst 2000 81 187 194 10869719 10.1016/S0165-1838(00)00137-5
Jordt SE McKemy DD Julius D Lessons from peppers and peppermint: the molecular logic of thermosensation Curr Opin Neurobiol 2003 13 487 492 12965298 10.1016/S0959-4388(03)00101-6
Fundin BT Pfaller K Rice FL Different distributions of the sensory and autonomic innervation among the microvasculature of the rat mystacial pad J Comp Neurol 1997 389 545 568 9421138 10.1002/(SICI)1096-9861(19971229)389:4<545::AID-CNE1>3.0.CO;2-0
Craig AD A new view of pain as a homeostatic emotion Trends Neurosci 2003 26 303 307 12798599 10.1016/S0166-2236(03)00123-1
Graven-Nielsen T Mense S The peripheral apparatus of muscle pain: evidence from animal and human studies Clin J Pain 2001 17 2 10 11289084 10.1097/00002508-200103000-00002
Light AR Perl ER Unmyelinated afferent fibers are not only for pain anymore J Comp Neurol 2003 461 137 139 12724832 10.1002/cne.10691
Kaufman MP Hayes SG The exercise pressor reflex Clin Auton Res 2002 12 429 439 12598947 10.1007/s10286-002-0059-1
Pan HL Chen SR Myocardial ischemia recruits mechanically insensitive cardiac sympathetic afferents in cats J Neurophysiol 2002 87 660 668 11826035
Lawson SN Scott SA Morphological and Biochemical Cell Types of Sensory Neurons Sensory Neurons: Diversity, Development, and Plasticity 1992 New York , Oxford University Press
Willis WDJ Dorsal root potentials and dorsal root reflexes: a double-edged sword Exp Brain Res 1999 124 395 421 10090653 10.1007/s002210050637
Sann H Pierau FK Efferent functions of C-fiber nociceptors Z Rheumatol 1998 57 Suppl 2 8 13 10025074 10.1007/s003930050226
Cannon KE Nalwalk JW Stadel R Ge P Lawson D Silos-Santiago I Hough LB Activation of spinal histamine H3 receptors inhibits mechanical nociception Eur J Pharmacol 2003 470 139 147 12798951 10.1016/S0014-2999(03)01737-0
Olson TH Riedl MS Vulchanova L Ortiz-Gonzalez XR Elde R An acid sensing ion channel (ASIC) localizes to small primary afferent neurons in rats Neuroreport 1998 9 1109 1113 9601677
Benson CJ Xie J Wemmie JA Price MP Henss JM Welsh MJ Snyder PM Heteromultimers of DEG/ENaC subunits form H+-gated channels in mouse sensory neurons Proc Natl Acad Sci U S A 2002 99 2338 2343 11854527 10.1073/pnas.032678399
Fundin BT Arvidsson J Aldskogius H Johansson O Rice SN Rice FL Comprehensive immunofluorescence and lectin binding analysis of intervibrissal fur innervation in the mystacial pad of the rat J Comp Neurol 1997 385 185 206 9268123 10.1002/(SICI)1096-9861(19970825)385:2<185::AID-CNE2>3.0.CO;2-Y
Rice FL Fundin BT Arvidsson J Aldskogius H Johansson O Comprehensive immunofluorescence and lectin binding analysis of vibrissal follicle sinus complex innervation in the mystacial pad of the rat J Comp Neurol 1997 385 149 184 9268122 10.1002/(SICI)1096-9861(19970825)385:2<149::AID-CNE1>3.0.CO;2-1
Park TJ Comer C Carol A Lu Y Hong HS Rice FL Somatosensory organization and behavior in naked mole-rats: II. Peripheral structures, innervation, and selective lack of neuropeptides associated with thermoregulation and pain J Comp Neurol 2003 465 104 120 12926019 10.1002/cne.10824
Pare M Elde R Mazurkiewicz JE Smith AM Rice FL The Meissner corpuscle revised: a multiafferented mechanoreceptor with nociceptor immunochemical properties J Neurosci 2001 21 7236 7246 11549734
Honig MG Hume RI Dil and diO: versatile fluorescent dyes for neuronal labelling and pathway tracing Trends Neurosci 1989 12 333 5, 340-1 2480673 10.1016/0166-2236(89)90040-4
Eckert SP Taddese A McCleskey EW Isolation and culture of rat sensory neurons having distinct sensory modalities J Neurosci Methods 1997 77 183 190 9489896 10.1016/S0165-0270(97)00125-8
Wessendorf MW Fluoro-Gold: composition, and mechanism of uptake Brain Res 1991 553 135 148 1933270 10.1016/0006-8993(91)90241-M
|
16305749
|
PMC1308857
|
CC BY
|
2021-01-04 16:40:08
|
no
|
Mol Pain. 2005 Nov 23; 1:35
|
utf-8
|
Mol Pain
| 2,005 |
10.1186/1744-8069-1-35
|
oa_comm
|
==== Front
Nutr JNutrition Journal1475-2891BioMed Central London 1475-2891-4-321628588310.1186/1475-2891-4-32ResearchSerum concentration of Selenium in healthy individuals living in Tehran Safaralizadeh R [email protected] GA [email protected] Z [email protected] M [email protected] A [email protected] S [email protected] Immunology, Asthma & Allergy Research Institute, Children Medical Center, Tehran University of Medical Sciences, Tehran, P.O. Box: 14185-938, I.R.Iran2005 14 11 2005 4 32 32 8 6 2005 14 11 2005 Copyright © 2005 Safaralizadeh et al; licensee BioMed Central Ltd.2005Safaralizadeh et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Objective
To investigate whether daily diet provides adequate selenium intake in healthy men and women living in Tehran, Iran.
Method
Serum level of selenium was determined in 184 healthy individuals of both genders. The samples were divided into two age groups, adults and children, for analysis. The serum level of selenium was determined using hydride generation and flame atomic absorption spectroscopy.
Results
The mean and standard deviation of serum selenium levels in children (1–16 years) was 84.3 ± 11 μg/l and there was no significant difference between genders in this group. In adults (older than 16 years) the mean serum selenium level was 100.6 ± 13 SD μg/l; among women the mean was 93.9 ± 14 SD μg/l and among men it was 102.2 ± 12 SD μg/l. The mean selenium level in men was higher than in women and data analysis showed a significant difference between them (p < 0.005). There was also a positive correlation between higher selenium serum concentration and age in men (P < 0.001). Daily intake of selenium in men and women was calculated to be 67 μg and 62.1 μg respectively.
Conclusion
Our results show that the serum concentration of selenium in an Iranian population is similar to other nationalities in the Middle East, particularly Saudi Arabia.
seleniumserumTehran inhabitants
==== Body
Introduction
Selenium (Se) plays a key role in the maintenance of normal health in human populations [1]. The cellular biochemistry of Selenium involves the expression of a variety of selenoproteins. Selenium is part of the active site of glutathione peroxidase (GSH-Px), an antioxidant enzyme [2]. It has been demonstrated that, when taken as a supplement, Selenium modulates the cellular response to oxidative stress, inducing a faster restoration of the endogenous antioxidative defense system against the production of reactive oxygen species [3]. Glutathion peroxidase controls the interacellular level of hydrogen peroxide, reducing the formation of reactive oxygen species that can induce lipid peroxidations with consequent damage to the cellular membranes [4]. Epidemiological studies suggest a low intake of Selenium might predispose an individual to an increased incidence of cardiovascular disorders [5]. There is increasing evidence that selenium deficiency may have several serious short- and long-term medical implications, including impaired immune response, or even cancer [6]. An experimental study has shown that an increase in selenium level is associated with decreased cancer mortality [7]. The recommended dietary allowance of Selenium in the USA is 55 μg/day for women and 70 μg/day for men. In some regions of the world such as Finland, New Zealand, the East coast of the United States of America and China the content of Selenium in soil is remarkably low [4]. Therefore Selenium levels in the serum of populations throughout the world vary from 41.7 μg/l in Finland to 158.2 μg/l in Canada [8]. There is currently no information of selenium intake and serum levels in the Iranian population. The aim of our study was therefore to evaluate serum levels of Selenium to find out whether daily diet provides adequate selenium intake to maintain the health of men and women living in Tehran.
Materials and methods
1. Subject selection
Serum samples were collected from 184 random inhabitants of Tehran. An informed consent was acquired, according to the guidelines from the Tehran university research ethic committee. Some of the samples were collected from excessive serum residues of blood taken from children who were referred for routine laboratory check up at the Children's Medical Center. All the samples were tested to rule out HIV, HBV and HCV contamination. Blood samples were left to coagulate spontaneously.
2. Determination of serum selenium
Blood samples (3 ml) were centrifuged at 3000 × g for 5 minutes. The clean serum was stored at -70°C until the time of analysis. All glassware and bottles used for the isolation of serum and for analysis were previously soaked in diluted nitric acid (10%) and rinsed thoroughly with deionized water. This procedure was followed in order to exclude the possibility of contamination with trace elements. Serum (500 μl) was aliquoted into a vessel-tube for mineralization with 3 ml of HNO3/HCLO4 (4:1 v/v). The temperature of this mixture was slowly increased to 175°C until fumes of HCLO4 appeared. The mixture was then heated according to the following (temperature/time) scheme: 175°C/60 min, 200°C/60 min and finally 250°C for 60 min. The mixture was then left to cool down to room temperature. HCL 6 N (10 ml) was added and heated to 170°C for 30 min to reduce the Se (VI) to Se (IV). After cooling to room temperature, Se concentration was determined using the hydride generation atomic absorption spectrometry (Atomic absorption spectrometer Shimadzu, AA-680). Sodium borohydride solution (3 g NaBH4, 1 g NaOH in 100 ml of mili-Q water) was used as a reducing agent. All samples and standards were analysed in duplicate. The accuracy of the procedure was evaluated by analyzing commercially available samples of lyophilized human serum seronorm™ trace element serum, level 1, M10181 indicating a recommended value of 81 μg/l, and seronorm™ trace element serum, level 2, NO0371 indicating a recommended value of 136 μg/l.
3. Statistical analysis
Kolmogorov-Smirnov tests were carried out for normal distribution. The reference range for serum selenium was determined as the 95% confidence interval (CI) of means. Differences in selenium concentration between the male and female populations were analyzed with the Mann-Whitney U-test. P-values of less than 0.05 were considered significant.
Results
The studied individuals were all healthy, non-smokers. Volunteer medical history and physical examination ruled out the presence of current disease in the studied individuals. None of the individuals showed any digestive symptoms indicative of nutrient malabsorption. The mean and standard deviation of the individual selenium levels in children (age 1–16) was 84.2 ± 11 μg/l, which was 85.1 ± 10.8 μg/l among females and 83.7 ± 11.2 μg/l among males. The mean serum selenium level in adults (over 16 years) was 100.6 ± 12.9 μg/l, for adult women the mean was 93.9 ± 13.6 μg/l and for adult men was 102.1 ± 12.3 μg/l (Table 1). The accuracy and precision of the methods used for selenium analysis are summarized in Table 2.
Table 1 Age and sex of the studied individuals with corresponding selenium levels (μg/l)
Age Sex n Min Max Mean ± SD Reference Range (95% CI)
1–16 Y F 22 66 105 85.0 ± 10.8 63–106
M 32 58 105 83.7 ± 11.1 63–106
Over16 Y F 24 74 125 93.9 ± 13.6 67–121
M 106 75 134 102.1 ± 12.2 79–126
Y: Year, M: Male, F: Female, SD: Standard Deviation
P-value less than 0.05 is statistically significant
Table 2 Sensitivity and precision of the assay
Material Concentration Accuracy (%) Precision R.S.D(%)
Certified Measured
Seronorm™ Trace Element Serum (Level 1, M10181) 81 ± 1.5 80.4 ± 2.7 99.3 3.36
Seronorm™ Trace Element Serum (Level 2, NO0371) 136 ± 4.5 135 ± 3.8 99.2 2.81
R.S.D: relative standard deviation
Discussion
Selenium is an essential mineral in human nutrition. Natural selenium present in the diet of humans is in the form of organic seleno-proteins such as selenomethionine and seleno-cysteine. Foods such as fish and whole grain cereals are especially rich in organic selenium compounds [11]. Selenium in cereals is primarily in the form of selenomethionine. This naturally occurring amino acid is the most important nutritional form of selenium.
Deficiencies of selenium contribute to the prevalence and severity of iodine deficiency disorders which are the most important and well-known global nutritional problems, primarily in less developed countries [12]. Iodine deficiency in childhood impairs neuromotor and intellectual development, with an average reduction in the intelligence quotient of 10 points [13]. Selenium is required in thyroid metabolism, converting inactive thyroid hormone into active thyroid hormone [14]. It has been shown that in goitrous children who are both Se and iodine deficient, major Se deficiency partially blunts thyroid response to iodine supplementation [15]. The mean serum Se level for healthy children (age 1–16) observed in this study was 84.2 ± 11 μg/l with no significant difference between sexes (p = 0.659). Table 3 shows the mean serum Se level of Iranian children compared to children from different countries.
Table 3 Comparison of the mean serum levels of selenium, among children from different countries (as mentioned previously [10])
Country Overall mean (μg/l) Age range (years)
Austria 48 1–15
Finland 58.5 1–15
Belgium 60 1–15
Germany 65.5 1–18
France 67 3–16
England 74.2 2–16
Iran 84.3 1–16
Japan 84.5 1–15
Slovenia 85 1–13
Turkey 89 1–16
US 106 1–18
Canada 126 1–9
The mean serum Se level for adults observed in this study was 100.6 ± 12.8 μg/l, which was similar to the one reported in a survey in Saudi Arabia [16]. In the Nutritional Prevention of Cancer (NPC) Trial [17], a Selenium level of 80 ng/mL is considered the minimum level of plasma selenium necessary in the bloodstream for maximum production of selenoproteins (glutathione peroxidases, thioredoxin reductase, etc.).
Our results show that in adults there is a significant difference between men and women (p < 0.005) with a higher concentration of selenium in men. This suggests a sex-linked hormonal influence over serum level of selenium. It has previously been shown that selenium is essential for spermatogenesis [18]. This trace element is present in the protein of the capsule surrounding the sperm mitochondria and may have a structural function [19]. Our data also show a positive correlation between a higher concentration of selenium in serum and age in men (P < 0.001).
Table 4 summaries the selenium serum levels in the Iranian population in comparison with different countries. It is higher than levels calculated for Finland and other countries where soil is poor in selenium content. By using the medium correlation factor (1.51) as introduced by Navarro et al., to estimate the daily intake of nutrients, the daily intake of Se was calculated as 62.19 μg in the female and 67 μg in the male populations [20].
Table 4 Comparison of the mean serum levels of selenium, among adults from different countries (as mentioned previously [9, 11])
Country Overall mean (μg/l)
Finland (Helsinki) 41.7
New Zealand (Dunedin) 47.2
Brazil (Rio de Janeiro) 73.2
W. Germany (Mainz) 81.1
Sweden (lund) 85.0
Italy (Rome) 89.8
Japan (Hiroshima) 97.6
Iran (Tehran) 100.3
Saudi Arabia 102.5
US (Mort.Gr.) 110.2
England (Southampton) 115.7
Canada (Toronto) 158.3
Considering the American RDA, which recommends a daily Se intake of 50 μg for women and 70 μg for men, it seems that the normal Iranian diet has an adequate content of selenium for both genders.
Tehran, the capital of Iran, is located in the north of the country and has a population of approximately fifteen million people, representing a large proportion of the country's total population, estimated to be seventy five million people. The diversity of nourishment sources, regional variation and different ethnic diets makes it difficult to extend these results to the whole population.
Acknowledgements
This study was supported in part by a grant from the Tehran University of Medical Sciences. The authors would like to acknowledge Dr. Verity A. Cadd at Oxford University for kind revision of the manuscript.
==== Refs
Savarino L Granchi D Ciapetti G Cenni E Ravaglia G Forti P Maioli F Mattioli R Serum concentrations of zinc and selenium in elderly people: results in healthy nonagenarians/centenarians Exp Gerontol 2001 36 327 39 11226746 10.1016/S0531-5565(00)00218-7
Rotruck JT Pope AL Ganther HE Swanson AB Hafeman DG Hoekstra WG Selenium: biochemical role as a component of glutathione peroxidase Science 1973 179 588 90 4686466
Jozanov-Stankov O Demajo M Djujic I Mandic M Selenium intake as a modulator of responsiveness to oxidative stress J Environ Pathol Toxicol Oncol 1998 17 251 7 9726798
Burk RF Levander OA Shils ME, Olson J, Shike M Selenium Modern nutrition in health and disease 1999 9 Lippincott Williams & Wilkins, Baltimore 265 276
Salonen JT Alfthan G Huttunen JK Pikkarainen J Puska P Association between cardiovascular death and myocardial infarction and serum selenium in a matched-pair longitudinal study Lancet 1982 2 175 9 6123886 10.1016/S0140-6736(82)91028-5
Kohrle J Brigelius-Flohe R Bock A Gartner R Meyer O Flohe L Selenium in biology: facts and medical perspectives Biol Chem 2000 381 849 64 11076017 10.1515/BC.2000.107
Gupta S Narang R Krishnaswami K Yadav S Plasma selenium level in cancer patients Indian J Cancer 1994 31 192 7 8557298
Lockitch G Selenium: clinical significance and analytical concepts Crit Rev Clin Lab Sci 1989 27 483 541 2690856
Muntau AC Streiter M Kappler M Roschinger W Schmid I Rehnert A Schramel P Roscher AA Age-related reference values for serum selenium concentrations in infants and children Clin Chem 2002 48 555 60 11861447
da Cunha S Filho FM Antelo DS de Souza MM Serum sample levels of selenium and copper in healthy volunteers living in Rio de Janeiro city Sci Total Environ 2003 301 51 4 12493184 10.1016/S0048-9697(02)00290-5
Schrauzer GN Katz RN Reductive dechlorination and degradation of mirex and kepone with Vitamin B12 Bioinorg Chem 1978 9 123 43 81074 10.1016/S0006-3061(00)80285-9
Vanderpas JB Contempre B Duale NL Goossens W Bebe N Thorpe R Ntambue K Dumont J Thilly CH Diplock AT Iodine and selenium deficiency associated with cretinism in northern Zaire Am J Clin Nutr 1990 52 1087 93 2239787
DeLong GR Leslie PW Wang SH Jiang XM Zhang ML Rakeman M Jiang JY Ma T Cao XY Effect on infant mortality of iodination of irrigation water in a severely iodine-deficient area of China Lancet 1997 350 771 3 9297997 10.1016/S0140-6736(96)12365-5
Napolitano G Bonomini M Bomba G Bucci I Todisco V Albertazzi A Monaco F Thyroid function and plasma selenium in chronic uremic patients on hemodialysis treatment Biol Trace Elem Res 1996 55 221 30 9096850
Zimmermann MB Adou P Toressani T Zeder C Hurrel RF Effect of oral iodized oil on thyroid size and thyroid hormone metabolism in children with concurrent selenium and iodine deficiency Eur J Clin Nutr 2000 54 209 13 10713742 10.1038/sj.ejcn.1600921
Raines DA Kinsara AJ Eid Fawzy M Vasudevan S Mohamed GE Legayada ES Al-Rawithi S El-Yazigi A Plasma and urinary selenium in Saudi Arabian patients with dilated cardiomyopathy Biol Trace Elem Res 1999 69 59 68 10383099
Duffield-Lillico AJ Reid ME Turnbull BW Combs GF JrSlate EH Fischbach LA Marshall JR Clark LC Baseline characteristics and the effect of selenium supplementation on cancer incidence in a randomized clinical trial: a summary report of the Nutritional Prevention of Cancer Trial Cancer Epidemiol Biomarkers Prev 2002 11 630 639 12101110
Wu SH Oldfield JE Whanger PD Weswig PH Effect of selenium, vitamin E and antioxidants on testicular function in rats Biol Reprod 1973 8 625 9 4736545
Hansen JC Deguchi Y Selenium and fertility in animals and man – a review Acta Vet Scand 1996 37 19 30 8659343
Navarro M Lopez H Ruiz ML Gonzalez S Perez V Lopez MC Determination of selenium in serum by hydride generation atomic absorption spectrometry for calculation of daily dietary intake Sci Total Environ 1995 175 245 52 8578307 10.1016/0048-9697(95)04859-6
|
16285883
|
PMC1308858
|
CC BY
|
2021-01-04 16:39:30
|
no
|
Nutr J. 2005 Nov 14; 4:32
|
utf-8
|
Nutr J
| 2,005 |
10.1186/1475-2891-4-32
|
oa_comm
|
==== Front
Plant MethodsPlant Methods1746-4811BioMed Central London 1746-4811-1-101628008310.1186/1746-4811-1-10MethodologyUsing genomic DNA-based probe-selection to improve the sensitivity of high-density oligonucleotide arrays when applied to heterologous species Hammond John P [email protected] Martin R [email protected] David J [email protected] Janet [email protected] Zoe F [email protected] Henrik J [email protected] Philip J [email protected] Sean T [email protected] Nottingham Arabidopsis Stock Centre, School of Biosciences, University of Nottingham, Sutton Bonington Campus, Loughborough, LE12 5RD, UK2 Plant Sciences Division, School of Biosciences, University of Nottingham, Sutton Bonington Campus, Loughborough, LE12 5RD, UK3 Warwick HRI, University of Warwick, Wellesbourne, Warwick, CV35 9EF, UK2005 9 11 2005 1 10 10 3 8 2005 9 11 2005 Copyright © 2005 Hammond et al; licensee BioMed Central Ltd.2005Hammond et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
High-density oligonucleotide (oligo) arrays are a powerful tool for transcript profiling. Arrays based on GeneChip® technology are amongst the most widely used, although GeneChip® arrays are currently available for only a small number of plant and animal species. Thus, we have developed a method to improve the sensitivity of high-density oligonucleotide arrays when applied to heterologous species and tested the method by analysing the transcriptome of Brassica oleracea L., a species for which no GeneChip® array is available, using a GeneChip® array designed for Arabidopsis thaliana (L.) Heynh. Genomic DNA from B. oleracea was labelled and hybridised to the ATH1-121501 GeneChip® array. Arabidopsis thaliana probe-pairs that hybridised to the B. oleracea genomic DNA on the basis of the perfect-match (PM) probe signal were then selected for subsequent B. oleracea transcriptome analysis using a .cel file parser script to generate probe mask files. The transcriptional response of B. oleracea to a mineral nutrient (phosphorus; P) stress was quantified using probe mask files generated for a wide range of gDNA hybridisation intensity thresholds. An example probe mask file generated with a gDNA hybridisation intensity threshold of 400 removed > 68 % of the available PM probes from the analysis but retained >96 % of available A. thaliana probe-sets. Ninety-nine of these genes were then identified as significantly regulated under P stress in B. oleracea, including the homologues of P stress responsive genes in A. thaliana. Increasing the gDNA hybridisation intensity thresholds up to 500 for probe-selection increased the sensitivity of the GeneChip® array to detect regulation of gene expression in B. oleracea under P stress by up to 13-fold. Our open-source software to create probe mask files is freely available and may be used to facilitate transcriptomic analyses of a wide range of plant and animal species in the absence of custom arrays.
Arabidopsis thalianaBrassica oleraceaBrassicaceaecross-speciesmicroarrayoligonucleotidephosphatephosphorusplant nutritionprobe maskingtranscriptomics
==== Body
Introduction
High-density oligonucleotide (oligo) arrays are a powerful and widely used tool for large-scale gene-expression profiling [1]. One extensively validated oligo array type is available as commercial GeneChip® technology (Affymetrix, Santa Clara, USA). GeneChip® arrays use probe-sets rather than single oligos per gene, comprising of between 11 and 20 probe-pairs to quantify abundance for each transcript. Each probe-pair consists of a perfect-match (PM) and a mismatch (MM) probe. The PM probe is a 25-base sequence complementary to the target transcript, whilst the MM probe is identical to the PM probe with the exception of a single mismatch at the 13th base. Transcript abundance can be calculated either from extrapolated hybridisation differences between PM and MM probes across a probe-set, or simply by using the PM data depending on the analysis used. GeneChip® arrays provide reproducible, accurate data at high throughput rates which can be easily stored and compared across experiments [2-4]. An example of the wide-scale adoption of GeneChip® technology can be seen within the plant sciences research community, where data from thousands of GeneChip® arrays under large numbers of experimental challenges on the model plant Arabidopsis thaliana (L.) Heynh. are publicly-available [4]. Unfortunately, at present, GeneChip® arrays are available for only a few species of eukaryotes. Thus, in contrast to A. thaliana, the transcriptomes of most agriculturally- or ecologically-important plant species are less extensively studied, since extensive sequence information and the fabrication of expensive custom arrays would be required before experimentation can begin.
One solution to this problem is to use GeneChip® arrays designed for model organisms to study closely related species [5-12]. Whilst this strategy is feasible, studies conducted to date have not accounted for the probability that inefficient hybridisation of transcripts from the target species to GeneChip® probes designed to the model species will attenuate the overall signal calculated across a probe-set [13]. For example, Ji et al. [13] describe a scenario where expression data analysed in Microarray Analysis Suite (MAS Version 5.0; Affymetrix), is based on the calculation that the weight carried by a probe within a probe-set is inversely related to its distance from the mean value of all probes within the probe-set. Thus, probes not generating signals due to sequence polymorphisms with the target organism will reduce the quality of information available from experiments using GeneChip® arrays designed for a model species to monitor the transcriptome of a closely related species. Ji et al. [13] thus adopted an RNA-based probe-selection system, to study non-human mammalian transcriptomes with human GeneChip® arrays, using probe mask files to exclude probes which hybridised weakly to their target transcript. However, this technique reduced the number of probe-sets available for transcriptome analysis by biasing the analysis towards the measurement of abundant transcripts.
To address the potential problems of using GeneChip® arrays designed for one species to monitor the transcriptome of a closely related species, we have developed a novel technique to improve the sensitivity of high-density oligonucleotide arrays when applied to heterologous species. In contrast to Ji et al. [13], this technique is based on selecting probe-pairs based on the hybridisation efficiency of the PM oligonucleotide probe with genomic DNA from a target species for which the GeneChip® was not originally designed. We initially used the A. thaliana ATH1-121501 GeneChip® array to study the transcriptome of Brassica oleracea L., which represents several important crops, including cabbage, kale, broccoli, Brussels sprout, cauliflower and kohlrabi, and half of the B. napus (oil-seed rape) genome. Brassica oleracea was chosen because the ancestral lineage of Arabidopsis and Brassica (family Brassicaceae) diverged only 12 to 19 million years ago [14]. The effects of genomic DNA-based probe-selection on estimates of transcriptional regulation in B. oleracea were quantified under an imposed physiological stress of phosphorus (P) deprivation. Probe mask files increased sensitivity of the ATH1-121501 GeneChip® array when detecting regulation of gene expression in B. oleracea under P stress by up to 13-fold. The response of B. oleracea to P stress was compared with that of A. thaliana [15-17]. The transcriptional response of plants to P stress can provide insights into nutrient signal pathways [18] and can inform efforts to improve fertiliser-use efficiency of crop production systems through breeding and nutritional diagnostics [16].
Results and Discussion
Genomic-DNA hybridisation and probe-selection
We used A. thaliana ATH1-121501 GeneChip® arrays to study the transcriptome of B. oleracea. Sequence polymorphisms between the two species are likely to result in an underestimate of transcript abundance if all probes are used within individual probe-sets [13]. Therefore, we selected subsets of probe-pairs from each probe-set on the A. thaliana GeneChip® array based on the hybridisation efficiency of genomic DNA from B. oleracea to homologous A. thaliana GeneChip® array PM probes.
Genomic DNA from B. oleracea was biotin-labelled and hybridised to the A. thaliana ATH1-121501 GeneChip® array. Probe-sets were selected for subsequent transcriptome analyses if the probe-set was represented by PM probes with gDNA hybridisation intensities above a set threshold. Selection was performed using a .cel file parser script written in the Perl programming language (Xspecies Version 1.1, available at ). The method was optimised empirically by generating 13 probe mask files with gDNA hybridisation intensity thresholds ranging from 0 (i.e. no probe-selection) to 1000. These 13 probe mask files were assessed in turn to investigate the quantification of the transcriptional response of B. oleracea to an imposed mineral nutrient (phosphorus; P) stress.
Arabidopsis thaliana PM probes hybridised extensively to the B. oleracea genomic DNA (Figure 1). When the gDNA hybridisation intensity threshold was increased from 0 to 1000 during probe mask file generation, PM probe retention in the probe mask files decreased rapidly. However, the retention of whole probe-sets, representing transcripts, were less sensitive to increases in gDNA hybridisation intensities during probe mask file generation, since only a minimum of one PM probe was required to retain a probe-set. For example, although the probe mask file generated using a gDNA hybridisation intensity threshold of 300 masked > 50 % of all PM probes, 98.8 % of available A. thaliana probe-sets were retained.
Figure 1 Number of Arabidopsis thaliana probe-pairs and probe-sets from the ATH1-121501 GeneChip® array used to study the transcriptome of Brassica oleracea var. alboglabra cv. A12DHd as a function of the gDNA hybridisation intensity thresholds used to generate the probe mask files. Filled circles are scaled to the left-hand y-axis (i.e. probe-sets used in probe mask files) and unfilled circles are scaled to the right-hand y-axis (i.e. probe-pairs used in probe mask files). Data were obtained by hybridising genomic DNA from B. oleracea to the A. thaliana ATH1-121501 GeneChip® array.
Testing DNA probe-selection: transcript abundance in control Brassica oleracea
Brassica oleracea were grown hydroponically using physiological techniques described elsewhere [15]. Control plants were supplied with a full nutrient solution throughout the experiment; treated (P-starved) plants were supplied with a nutrient solution containing no P for 100 h. Transcriptional responses of B. oleracea to phosphate stress were determined by challenging A. thaliana ATH1-121501 GeneChip® arrays with total RNA extracted from control and treated plants. Following scanning of the GeneChip® arrays, raw cell intensity data files (.cel files) were loaded into GeneSpring (Silicon Genetics, CA, USA) using 13 probe mask files. For each of the 13 probe-selection conditions, data were prenormalised using Robust Multichip Average (RMA) algorithms [19]. Subsequently, data from the eight GeneChip® arrays (four control and four P-starved samples) were treated as 13 individual 'experimental scenarios' to evaluate probe-selection stringency. Within each 'scenario', data were further standardised within GeneSpring; the signal value from each replicate P-starved plant was standardised to its corresponding control sample to give an expression ratio for each gene.
The use of probe masks increased estimates of transcript abundance in B. oleracea in both control and P-starved samples (data shown for control samples; Figure 2a–e). The probe mask file generated with a gDNA hybridisation intensity threshold of 50 did not affect estimates of transcript abundance; only nine of 250,206 available PM probes were excluded from this probe mask file. Thus, all probe-sets were retained for subsequent transcriptome analysis (Figures 1, 2a). However, probe mask files generated with gDNA hybridisation intensity thresholds of 100 and above increased estimates of transcript abundance compared to using no probe-selection (Figure 2b–d). Estimates of transcript abundance in B. oleracea increased by over 2.5-fold when probe masks were used (Figure 2e). Increasing the gDNA hybridisation intensity threshold during probe mask file generation also increased the coefficient of variation (CV) calculated for ranked transcripts (probe-sets) across the four replicate control arrays (Figure 2f). Thus, if a stringent probe-selection criterion is to be used to analyse experimental transcriptome data, it may be appropriate to increase biological replication. However, the CV of control arrays were minimally affected using probe mask files generated at gDNA hybridisation intensity thresholds below 500 (Figure 2f).
Figure 2 (a) – (d) Probe-set signals of genes in control (P-replete) Brassica oleracea var. alboglabra cv. A12DHd estimated following probe-selection compared to probe-set signals estimated without probe-selection. Data are presented using probe mask files generated with gDNA hybridisation intensity thresholds of 50, 100, 200 and 400 respectively ((a) – (d)). Mean values (e) and ranked coefficient of variation (f) of probe-set signals of control (P-replete) B. oleracea as a function of the gDNA hybridisation intensity thresholds used to generate probe mask files for the transcriptome analysis. In (f), the gDNA hybridisation intensity threshold used to generate probe mask files is indicated by different coloured lines: red (gDNA hybridisation intensity threshold = 0), green (50), yellow (100), blue (150), pink (200), cyan (300), black (400), grey (500), dark red (600), dark green (700), dark pink (800), dark cyan (900), dark yellow (1000). In all panels, total RNA samples were extracted from the shoots of hydroponically-grown control (P-replete) B. oleracea (n = 4).
Testing DNA probe-selection: gene regulation under P stress in Brassica oleracea
Gene expression ratios were log-normally distributed when experiments were analysed using any of the 13 probe mask files (data not shown). For each experimental interpretation, fold-change differences in gene expression were calculated as the ratio of normalised means for each gene in the treated sample relative to the control sample. A one-way ANOVA, with a Benjamini and Hochberg False Discovery Rate (BH-FDR = 0.05) multiple testing correction (MTC) applied, was used to test if transcript abundance was significantly different between control and P-starved samples. Increasing the gDNA hybridisation intensity threshold from 0 to 500 for probe mask file generation increased the sensitivity of the ATH1-121501 GeneChip® array to detect regulation of gene expression in B. oleracea under P stress, in terms of significant > ± 1.3-fold differences in gene expression between P-starved and control samples (Figures 3, 4).
Figure 3 'Volcano' plots illustrating the log2 of the fold-changes (i.e. the ratio of means for each gene) and inverse significance (i.e. log10 of the reciprocal of the Benjamini and Hochberg False Discovery Rate multiple test corrected P-value derived from a one-way ANOVA with the Benjamini and Hochberg FDR multiple testing correction) in gene expression differences between control and P-starved Brassica oleracea var. alboglabra cv. A12DHd. Total RNA samples were extracted from control B. oleracea shoots and from the shoots of plants grown in the absence of P for 100 h (n = 4). (a) no probe-selection used during transcriptome analysis, (b), (c), (d) using probe mask files during transcriptome analysis, generated at gDNA hybridisation intensity thresholds of 200, 400 and 1000 respectively. Data in Additional file 1.
Figure 4 Gene regulation under P-starvation in Brassica oleracea var. alboglabra cv. A12DHd as a function of the gDNA hybridisation intensity threshold used to generate probe mask files for the transcriptome analysis. Total RNA samples were extracted from control B. oleracea shoots and from the shoots of plants grown in the absence of P for 100 h (n = 4). (a) genes significantly regulated under P starvation at Benjamini and Hochberg False Discovery Rate multiple test corrected (BH-FDR MTC) P < 0.05. (b) genes regulated > ± 1.3-fold under P starvation. (c) genes significantly regulated > ± 1.3-fold under P starvation (BH-FDR MTC P < 0.05). Data in Additional file 1.
Estimates of fold-differences in gene expression (P-starved samples versus control samples) increased using probe mask files generated with gDNA hybridisation intensity thresholds up to 500 (Figure 3, 4c). 'Volcano' plots summarise these effects (Figure 3a–c). In volcano plots, the y-axis represents the log10 of the reciprocal of the one-way ANOVA P-value corrected for multiple testing using the BH-FDR; the x-axis represents the log2 of the fold-change in gene expression. There was a 13-fold increase in the number of genes identified as differentially regulated > ± 1.3-fold under P stress using a probe mask file generated with a gDNA hybridisation intensity threshold of 500 (Figures 3, 4c). The use of probe mask files generated with higher gDNA hybridisation intensity thresholds (up to 1000) resulted in a substantial loss of probe-sets available for transcriptome analysis (Figures 1, 3d). However, the number of genes identified as differentially regulated > ± 1.3-fold was still much greater when using the probe mask file generated with a gDNA hybridisation intensity threshold of 1000 than when probe-selection was not used, even though only 7.3 % of the total number of PM probes on the GeneChip® array were used (Figure 4c). Estimates of the number of genes significantly differentially regulated under P stress (at BH-FDR MTC P < 0.05) increased when probe mask files were used (Figure 4a). For example, using a probe mask file generated with a gDNA hybridisation intensity threshold of 500, 111 genes were estimated as significantly differentially regulated, compared to only eight genes regulated using no probe-selection.
An optimal probe-selection strategy for transcriptional analysis of B. oleracea is to use a probe mask file generated at a gDNA hybridisation intensity threshold of 200 or greater. Although estimates of fold-differences in gene expression in P-starved versus control samples were greatest using probe mask files generated with gDNA hybridisation intensity thresholds of up to 500 (Figure 4c,d), there was a significant loss of available probe-sets for transcriptome analysis with gDNA hybridisation intensity thresholds > 500. Further, estimates of the number of significantly regulated genes declined using probe mask files generated at DNA hybridisation intensity thresholds > 500.
Biological significance of genes regulated under P stress in Brassica oleracea
Following probe-selection using a gDNA hybridisation intensity threshold of 400, 99 genes were identified as significantly differentially regulated in the shoots of B. oleracea following the withdrawal of P from the nutrient solution (BH-FDR MTC P < 0.05; Additional file 1). Of these genes, 39 had higher transcript abundance in P-starved samples than in control samples and 60 had lower transcript abundance in P-starved samples than in control samples. Data for all genes, analysed with all probe mask files generated at gDNA hybridisation intensity thresholds from 0 to 1000, are presented in Supplementary Table I.
Sequence polymorphisms between A. thaliana and B. oleracea are likely to result in the identification of homologous genes or alternative members of gene families in addition to possible gene orthologues. For this reason, we compared previously published studies on the response of A. thaliana to P starvation [15-17] at the level of individual gene and also at the level of the gene family and functional category. Several homologous genes responded similarly to P starvation in B. oleracea and A. thaliana. For example, SQD2 (At5g01220), which is involved in sulpholipid biosynthesis, increases its expression in responses to P starvation in A. thaliana [15-17,20-22]. A homologue of this gene had higher hybridisation signals in P-starved B. oleracea than in control samples. The signal values for SQD2, obtained using the probe mask file generated with a gDNA hybridisation intensity threshold of 400, were based on two probes, in contrast to 11 probes when no probe-selection was used. The use of probe-selection increased the significance and differences in transcript abundance between P-starved and control plants (SQD2, 1.60 ± 0.2, BH-FDR MTC P = 0.047; mean normalised signal ratio ± 1 S.D.) than in the absence of probe-selection (SQD2, 1.53 ± 0.58, BH-FDR MTC P = 0.382).
Ribonucleases, phosphatases and phosphodiesterases are involved in the recycling of P in plants during P starvation [16,23,24]. The hybridisation intensity of the ribonuclease RNS2 (At2g39780) was higher in P-starved samples compared to control samples, and is thought to be involved in the release of P from internal sources during P starvation [23]. The signal value for RNS2, obtained using the probe mask file generated with a gDNA hybridisation intensity threshold of 400, was based on six probes in contrast to 11 probes when no probe-selection was used. The use of probe-selection increased normalised signal ratios in P-starved plants compared to control plants to a greater amount (RNS2, 1.92 ± 0.43, BH-FDR MTC P = 0.06, mean normalised signal ratio ± 1 S.D.) than in the absence of probe-selection, where the up-regulation of the gene was not detected (RNS2, 1.07 ± 0.06, BH-FDR MTC P = 0.43).
Higher hybridisation signals in P-starved plants compared to control samples were also observed for genes belonging to gene families or functional groups which have previously been observed responding to P starvation in A. thaliana, although not significantly [15-17]. These include genes involved in the transport of phosphate and phosphate-containing compounds and carbohydrates, indicating a change in the internal economy of P use in response to P starvation (Additional file 1). For example, at a gDNA hybridisation intensity of 400, two sucrose-phosphate synthases, and a malic enzyme had higher hybridisation intensities for in P-starved compared to control samples, whilst lower hybridisation intensities for genes involved in alternative glycolosis reactions that conserve P during P starvation conditions [16] were noted. There are other notable similarities in the P starvation response of B. oleracea with published responses of A. thaliana to P starvation at the level of the gene family. These include transcription factors from the zinc finger, F-box, bHLH and myb families, genes involved in flavanoid biosynthesis and genes encoding proteins from the cytochrome P450, glycosyl hydrolase, and peroxidase families.
Conclusion
Since GeneChip® arrays are available for only a small number of species, we have developed a novel probe-selection method to study the transcriptome of a plant or animal species for which GeneChip® arrays are not available. Genomic DNA from B. oleracea was labelled and hybridised to the A. thaliana ATH1-121501 GeneChip® array. Perfect-match A. thaliana probes which hybridised to the B. oleracea genomic DNA above selected hybridisation intensities were selected for subsequent B. oleracea transcriptome analysis using probe mask files generated using a .cel file parser script. Software to create probe mask files is freely available and is designed to facilitate the analysis of the transcriptomes of a wide range of other species in the absence of custom arrays.
Probe-selection was tested by quantifying the transcriptional response of B. oleracea to a mineral nutrient (phosphorus; P) stress using probe mask files generated at gDNA hybridisation intensity thresholds ranging from 0 (i.e. no probe-selection) to 1000. A probe mask file generated at a gDNA hybridisation intensity threshold of 400 masked > 68 % of all of the A. thaliana probe-pairs from the subsequent transcriptome analysis whilst retaining 96.4 % of the total available probe-sets. Increasing the gDNA hybridisation intensity thresholds for probe mask file generation increased the sensitivity of the GeneChip® array to detect regulation of gene expression in B. oleracea under P stress by up to 13-fold. Ninety-nine genes were significantly regulated in the shoots of B. oleracea under P stress (BH-FDR MTC P < 0.05). Confirmatory analyses using quantitative PCR and reporter-genes fused to nutrient responsive promoters have already been conducted for several P stress regulated genes in A. thaliana [15,16] and detailed confirmatory and functional molecular analysis of these genes is now clearly required in B. oleracea to resolve the biological significance of the P stress response. Understanding P stress responses in plants is likely, (i) to hasten the development of more nutrient efficient varieties of crops, (ii) to improve our understanding of nutrient signalling pathways, and (iii) to yield rapid advances in nutritional diagnostics, in particular if unit costs for microarray analyses continue to decrease [16,18]. We appreciate that the resolution and clarity of the probe mask method will be influenced by gene duplication and will necessitate validation of the potential homologues found in the target species, however this caution is generally true in the design of any new microarray. Most array users are primarily interested in transcript differences between samples as clues to further investigation of a gene or gene family. This technique enables such differential expression indicators for genes from unusual species and additionally provides putative model organism gene homologue ontologies via the standard Affymetrix gene annotation. As a pragmatic method based on gDNA hybridisation the technique requires investigation of a range of thresholds to systematically test probe-selection at the bioinformatics level. Re-optimisation of gDNA hybridisation thresholds is particularly important for new species given that gDNA quality and origin may affect overall absolute hybridisation intensities per experiment but should not affect relative ranking of intensities between hybridisations of the same genome. The technique therefore allows an economical analysis of any RNA hybridisations performed, since gDNA-based probe-selection can be optimised at any time without repeating the RNA work. The method also requires minimal setup outlay represented by one representative gDNA hybridisation and some straightforward bioinformatics time, in contrast to the costs of manufacturing even a simple single oligo array. More importantly, however, gDNA-based probe-selection could facilitate transcriptional profiling of a wide range of plant and animal species of agricultural, ecological and evolutionary importance, even in the absence of available genomic information.
Materials and methods
Genomic DNA hybridisation and probe-selection
Probe-pairs from the A. thaliana ATH1-121501 GeneChip® array (Affymetrix, Santa Clara, CA, USA) were selected for transcriptome analysis of B. oleracea using a gDNA-based probe-selection strategy based on the hybridisation of gDNA to the PM probe. Total genomic DNA was extracted from 5 g of B. oleracea var. alboglabra cv. A12DHd leaf tissue using a DNeasy Plant mini kit (Qiagen Ltd, Crawley, UK). Genomic DNA was labelled using the Bioprime DNA labelling System (Invitrogen, Paisley, UK) and subsequently hybridised to Affymetrix ATH1-121501 GeneChip® arrays for 16 h at 45°C using standard Affymetrix hybridisation protocols. Subsequently, the GeneChip® array was scanned on an Affymetrix G2500A GeneArray scanner and a cell intensity file (.cel file) was generated using Microarray Analysis Suite (MAS Version 5.0; Affymetrix). This .cel file contained the gDNA hybridisation intensities between B. oleracea genomic DNA fragments and all A. thaliana probes. Probe-pairs from the .cel file were selected for subsequent transcriptome analysis using a .cel file parser script (Xspecies Version 1.1) written in the Perl programming language . The Perl script was designed to create probe mask (.cdf) files compatible with a range of microarray analysis software packages. A probe-set was selected when it was represented by one or more PM probe-pair(s) per probe-set (i.e. a minimum of 25 bp identical probe sequence to A. thaliana was required for subsequent transcriptome analysis of B. oleracea). There was no a priori restriction of a suitable gDNA hybridisation intensity threshold for probe mask file generation in a target species. Thus, the algorithm was designed to allow a user-specified gDNA hybridisation intensity threshold for probe mask file generation to be set. Files (cdf.) were therefore generated using a range of gDNA hybridisation intensity thresholds (from 0 to 1000). The Perl algorithm and B. oleracea DNA .cel files are freely available from along with similar files for other species. The Perl masking script and instructions for using the script are also available as Additional file 2 and Additional file 3.
Testing the gDNA-based probe-selection strategy: determining the transcriptional response of B. oleracea to phosphate stress
Brassica oleracea var. alboglabra cv. A12DHd were grown hydroponically from seed in a system described in Hammond et al., [15]. Control plants were supplied with nutrient solution containing all nutrients throughout the experiment whilst treated plants were supplied with a nutrient solution containing no P for 100 h. Replicate samples (each replicate comprising 8–10 individual plants) were harvested from control and treated plants, mid way through the photo-period, and snap frozen in liquid nitrogen. All plants had approximately 6–8 leaves at harvest. Three biological replicates and one technical replicate were used for transcriptome analysis.
Tissue samples, previously stored at -70°C, were placed in liquid nitrogen before grinding. To each sample, 1 ml of TRIzol reagent (Invitrogen) was added and total RNA was subsequently extracted as described previously [15]. Total RNA yield and purity were determined using an Agilent 2100 Bioanalyser (Agilent Technologies, Stockport, Cheshire, UK). Approximately 5 μg of total RNA was reverse transcribed at 42°C for 1 h to generate first strand cDNA using 100 pmol oligo dT(24) primer containing a 5'-T7 RNA polymerase promoter sequence, 50 mM Tris-HCl (pH 8.3), 75 mM KCl, 3 mM MgCl2, 10 mM dithiothreitol (DTT), 10 mM dNTPs and 200 units SuperScript II reverse transcriptase (Invitrogen Life Technologies). Following first strand cDNA synthesis, second strand cDNA was synthesised using 10 units of Escherichia coli polymerase I, 10 units of E. coli DNA ligase and 2 units of RNase H in a reaction containing 25 mM Tris-HCl (pH 7.5), 100 mM KCl, 5 mM MgCl2, 10 mM (NH4)SO4, 0.15 mM b-NAD+ and 10 mM dNTPs. The second strand synthesis reaction proceeded at 16°C for 2 h before 10 units of T4 DNA polymerase was added and the reaction allowed to proceeded for a further 5 minutes. The reaction was terminated by adding 0.5 M EDTA. Double stranded cDNA products were purified using the GeneChip® Sample Cleanup Module (Affymetrix). The synthesised cDNAs were in-vitro transcribed by T7 RNA polymerase (Enzo BioArray High Yield RNA Transcript Labelling Kit, Enzo Life Sciences Inc., Farmingdale, NY, USA) using biotinylated nucleotides to generate biotinylated complementary RNAs (cRNAs). The cRNAs were purified using the GeneChip® Sample Cleanup Module (Affymetrix). The cRNAs were then randomly fragmented at 94°C for 35 minutes in a buffer containing 40 mM Tris-acetate (pH 8.1), 100 mM potassium acetate, and 30 mM magnesium acetate to generate molecules of approximately 35 to 200 bp. Affymetrix A. thaliana ATH1-121501 GeneChip® arrays were hybridised with 15 μg of fragmented labelled cRNA for 16 h at 45°C as described in the Affymetrix Technical Analysis Manual. GeneChip® arrays were stained with Streptavidin-Phycoerythrin solution and scanned with an Affymetrix G2500A GeneArray scanner.
Microarray Analysis Suite (MAS Version 5.0; Affymetrix) was used to generate .cel files for each of the biological RNA replicates by scanning and computing summary intensities for each probe without the use of probe mask files. These .cel files were loaded into GeneSpring (Agilent Technologies) analysis software package using the Robust Multichip Average (RMA) pre-normalisation algorithm (Irizarry et al., 2003). During .cel file loading and pre-normalisation, .cel files were interpreted using either, (1) the A. thaliana .cdf file (i.e. with no probe-selection used), or (2) using .cdf files generated from the gDNA .cel file with gDNA hybridisation intensity thresholds from 50 to 1000. Following RMA pre-normalisation and masking of individual probes, further standardisations were applied to the probe-set raw signal value in GeneSpring. For each replicate array, each probe-set signal value from treated (P-starved) samples was standardised to the probe-set signal value of its corresponding control sample to give gene expression ratios between the two conditions. Genes with differential hybridisation intensities between P-starved and control samples were identified using a one-way ANOVA, with a Benjamini and Hochberg false discovery rate (0.05) multiple testing correction applied, to identify genes that were significantly differentially expressed between the two conditions. Genes with Benjamini and Hochberg FDR multiple test corrected P-values less than 0.05 were considered to be differentially regulated under P starvation.
Supplementary Material
Additional File 1
An archive of the mean probe-set signals of all genes in P-starved versus control (P-replete) Brassica oleracea var. alboglabra cv. A12DHd. Raw data were analysed with the RMA pre-normalisation algorithm using the ATH1-121501 probe mask file (no probe selection) or a custom probe mask file. The custom probe mask files were generated by hybridising gDNA from B. oleracea to Affymetrix A. thaliana ATH-121501 GeneChip® arrays and selecting probe-pairs in which B. oleracea gDNA hybridisation intensity values were greater than 50, 100, 150, 200, 300,...1000. Signal intensities are estimates from total RNA extracted from the shoots of hydroponically-grown plants (n = 4) under P-deficient (raw) and P-replete (control) conditions. The P-values are derived from a one-way ANOVA using a Benjamini and Hochberg False Discovery Rate (0.05) multiple testing correction.
Click here for file
Additional File 2
An archive containing all of the Perl scripts necessary in order to produce new CDF files for use in analysing cross-species hybridisation results.
Click here for file
Additional File 3
(CDF_filtering script instructions.doc : DOC file). An instruction document containing details for using the scripts contained within Additional file 2.
Click here for file
Acknowledgements
Seeds of Brassica oleracea var. alboglabra cv. A12DHd were kindly provided by Graham King (Warwick HRI, UK). J. Okyere (University of Nottingham, UK) and Santhoshi Bandla, Cheng Li (Department of Biostatistics, Harvard School of Public Health, US) provided helpful technical discussions in the early stages of the project. The work was supported financially by the UK Department for the Environment, Food and Rural Affairs (grants HH3501SFV and HH3504SPO), by the UK BBSRC Investigating Gene Function Arabidopsis award (IGF12422), and by a University of Nottingham New Lecturer's Award.
==== Refs
Lipshutz RJ Fodor SPA Gingeras TR Lockhart DJ High density synthetic oligonucleotide arrays Nature Genet 1999 21 20 24 9915496 10.1038/4447
Harmer SL Hogenesch LB Straume M Chang HS Han B Zhu T Wang X Kreps JA Kay SA Orchestrated transcription of key pathways in Arabidopsis by the circadian clock Science 2000 290 2110 2113 11118138 10.1126/science.290.5499.2110
Zhu T Wang X Large scale profiling of the Arabidopsis transcriptome Plant Physiol 2000 124 1472 1476 11115862 10.1104/pp.124.4.1472
Craigon DJ James N Okyere J Higgins J Jotham J May S NASCArrays: a repository for microarray data generated by NASC's transcriptomics service Nucleic Acids Res 2004 32 D575 D577 14681484 10.1093/nar/gkh133
Chismar JD Mondala T Fox HS Roberts E Langford D Masliah E Salomon DR Head SR Analysis of results variability from high-density oligonucleotide arrays comparing same-species and cross-species hybridisations Biotechniques 2002 33 516 524 12238761
Enard W Khaitovich P Klose J Zöllner S Heissig F Giavalisco P Nieselt-Struwe K Muchmore E Varki A Ravid R Doxiadis GM Bontrop RE Pääbo S Intra-and interspecific variation in primate gene expression patterns Science 2002 296 340 343 11951044 10.1126/science.1068996
Caceres M Lachuer J Zapala MA Redmond JC Kudo L Geschwind DH Lockhart DJ Preuss TM Barlow C Elevated gene expression levels distinguish human from non-human primate brains Proc Natl Acad Sci USA 2003 100 13030 13035 14557539 10.1073/pnas.2135499100
Higgins MA Berridge BR Mills BJ Schultze AE Gao H Searfoss GH Baker TK Ryan TP Gene Expression analysis of the acute phase response using a canine microarray Toxicol Sci 2003 74 470 484 12773757 10.1093/toxsci/kfg142
Becher M Talke IN Krall L Krämer U Cross-species microarray transcript profiling reveals high constitutive expression of metal homeostasis genes in shoots of the zinc hyperaccumulator Arabidopsis halleri Plant J 2004 37 251 268 14690509
Khaitovich P Weiss G Lachmann M Hellmann I Enard W Muetzel B Wirkner U Ansorge W Pääbo S A neutral model of transcriptome evolution PLo S Biol 2004 2 682 689
Uddin M Wildman DE Liu GZ Xu WB Johnson RM Hof PR Kapatos G Grossman LI Goodman M Sister grouping of chimpanzees and humans as revealed by genome-wide phylogenetic analysis of brain gene expression profiles Proc Natl Acad Sci USA 2004 101 2957 2962 14976249 10.1073/pnas.0308725100
Weber M Harada E Vess C v. Roepenack-Lahaye E Clemens S Comparative microarray analysis of Arabidopsis thaliana and Arabidopsis halleri roots identifies nicotianamine synthase, a ZIP transporter and other genes as potential metal hyperaccumulation factors Plant J 2004 37 269 281 14690510 10.1111/j.1365-313X.2003.02013.x
Ji W Zhou W Gregg K Yu N Davis S Davis S A method for cross-species gene expression analysis with high-density oligonucleotide arrays Nucleic Acids Res 2004 32 e93 doi:10.1093/nar/gnh084. 15247326 10.1093/nar/gnh084
AGI (The Arabidopsis Genome Initiative) Analysis of the genome sequence of the flowering plant Arabidopsis thaliana Nature 2000 408 796 815 11130711 10.1038/35048692
Hammond JP Bennett MJ Bowen HC Broadley MR Eastwood DC May ST Rahn C Swarup R Woolaway KE White PJ Changes in gene expression in Arabidopsis shoots during phosphate starvation and the potential for developing smart plants Plant Physiol 2003 132 578 596 12805589 10.1104/pp.103.020941
Hammond JP Broadley MR White PJ Genetic responses to phosphorus deficiency Ann Bot 2004 94 323 332 15292042 10.1093/aob/mch156
Wu P Ma L Hou X Wang M Wu Y Liu F Deng XW Phosphate starvation triggers distinct alterations of genome expression in arabidopsis roots and leaves Plant Physiol 2003 132 1260 1271 12857808 10.1104/pp.103.021022
Ticconi CA Abel S Short on phosphate: plant surveillance and countermeasures Trends in Plant Science 2004 9 548 555 15501180 10.1016/j.tplants.2004.09.003
Irizarry RA Hobbs B Collin F Beazer-Barclay YD Antonellis KJ Scherf U Speed TP Exploration, normalization, and summaries of high density oligonucleotide array probe level data Biostatistics 2003 4 249 264 12925520 10.1093/biostatistics/4.2.249
Essigmann B Güler S Narang RA Linke D Benning C Phosphate availability affects the thylakoid lipid composition and the expression of SQD1, a gene required for sulfolipid biosynthesis in Arabidopsis thaliana Proc Natl Acad Sci USA 1998 95 1950 1955 9465123 10.1073/pnas.95.4.1950
Yu B Xu C Benning C Arabidopsis disrupted in SQD2 encoding sulfolipid synthase is impaired in phosphate-limited growth Proc Natl Acad Sci USA 2002 99 5732 5737 11960029 10.1073/pnas.082696499
Kobayashi K Awai K Takamiya K-I Ohta H Arabidopsis type B monogalactosyldiacylglycerol synthase genes are expressed during pollen tube growth and induced by phosphate starvation Plant Physiol 2004 134 640 648 14730084 10.1104/pp.103.032656
Bariola PA Howard CJ Taylor CB Verburg MT Jaglan VD Green PJ The Arabidopsis ribonuclease gene RNS1 is tightly controlled in response to phosphate limitation Plant J 1994 6 673 685 8000425 10.1046/j.1365-313X.1994.6050673.x
van der Rest B Boisson A-M Gout E Bligny R Douce R Glycerophosphocholine metabolism in higher plant cells. Evidence of a new glyceryl-phosphodiester phosphodiesterase Plant Physiol 2003 130 244 255 12226504 10.1104/pp.003392
|
16280083
|
PMC1308859
|
CC BY
|
2021-01-04 16:37:28
|
no
|
Plant Methods. 2005 Nov 9; 1:10
|
utf-8
|
Plant Methods
| 2,005 |
10.1186/1746-4811-1-10
|
oa_comm
|
==== Front
Plant MethodsPlant Methods1746-4811BioMed Central London 1746-4811-1-111628751010.1186/1746-4811-1-11ReviewBlue-native PAGE in plants: a tool in analysis of protein-protein interactions Eubel Holger [email protected] Hans-Peter [email protected] A Harvey [email protected] ARC Centre of Excellence in Plant Energy Biology, University of Western Australia, 35 Stirling Hwy, Crawley 6009, Perth, Australia2 Abteilung Angewandte Genetik, Universität Hannover, Herrenhäuser Str. 2, 30419 Hannover, Germany2005 16 11 2005 1 11 11 5 9 2005 16 11 2005 Copyright © 2005 Eubel et al; licensee BioMed Central Ltd.2005Eubel et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Intact protein complexes can be separated by apparent molecular mass using a standard polyacrylamide gel electrophoresis system combining mild detergents and the dye Coomassie Blue. Referring to the blue coloured gel and the gentle method of solubilization yielding native and enzymatically active protein complexes, this technique has been named Blue-Native Polyacrylamide Gel-Electrophoresis (BN-PAGE). BN-PAGE has become the method of choice for the investigation of the respiratory protein complexes of the electron transfer chains of a range of organisms, including bacteria, yeasts, animals and plants. It allows the separation in two dimensions of extremely hydrophobic protein sets for analysis and also provides information on their native interactions. In this review we discuss the capabilities of BN-PAGE in proteomics and the wider investigation of protein:protein interactions with a focus on its use and potential in plant science.
Gel-based Proteomics2D-PAGEprotein complexeshydrophobic proteinsCoomassiesolubilization
==== Body
Introduction
With several plant genome sequencing projects already finished and several others nearing completion, the logical next step is the assessment of the protein products encoded by these genomes. The protein composition of different tissues, cell types and isolated cellular organelles from a range of plant species have been investigated to date making use of peptide mass spectrometry and pattern matching to genomic data. These studies have used a variety of approaches to separate proteins, but isoelectric focussing (IEF) followed by SDS-PAGE (typically termed a 2D gel) is the most common technique undertaken. This dominance is despite the fact that this technique is severely limited in its ability to display the very large complement of hydrophobic proteins from plants. Further, even if improvements in this standard 2D gel technique could further alleviate this hydrophobicity problem, a complete understanding of the processes taking place within cells will require much more than the simple identification of the individual polypeptides forming the proteome. Most cellular processes require the action of several enzymes, often each containing multiple subunits. To raise the efficiency, specificity and speed of metabolic pathways, these enzymes are often associated with each other, forming temporary or stable larger protein complexes. Knowledge of the composition and/or structure of these protein complexes will result in a much deeper understanding of metabolic pathways and cellular processes than protein identities alone are able to deliver.
There are many ways to investigate protein interactions, each with individual advantages and drawbacks. Many of the approaches commonly used today are investigative studies of the actual or possible interaction partner(s) of a particular protein of interest. Examples include yeast two-hybrid systems, immunoprecipitation studies with specific antibodies and the more recent use of TAP/FLAG pull-down assays, and FRET/BRET fluorescence studies in vivo. However, these approaches are not designed to provide a general overview of protein-protein interaction in a complex proteome of choice by a single experiment.
For an in-depth investigation of the protein complexes forming the respiratory chain of various organisms, Schägger et al. [1] developed a novel experimental strategy to investigate the individual components of this biochemical pathway. Through the combination of mild detergents and the dye Coomassie blue, substituting for the highly denaturating detergent sodium dodecyl sulphate (SDS), it was possible for the first time to separate intact respiratory protein complexes by electrophoresis. Referring to the blue coloured gel and the gentle method of solubilization yielding native and enzymatically active protein complexes, this technique has been named Blue-Native Polyacrylamide Gel-Electrophoresis (BN-PAGE). Over the last 10 years, BN-PAGE in combination with second dimension SDS-PAGE has been the method of choice for the investigation of the respiratory protein complexes of the electron transfer chains of a range of organisms, including bacteria, yeasts, animals and plants. This technique allows the separation in two dimensions of extremely hydrophobic protein sets for analysis and also provides information on their native interactions. There is now a growing number of publications employing this method for the investigation of other hydrophobic and hydrophilic high molecular weight protein complexes in different organisms. Recently, with the introduction of even more sophisticated solubilization methods, BN-PAGE has been used to detect specific interactions between large protein complexes that has led to the discovery of so-called 'supercomplexes'.
In this review we discuss the capabilities of BN-PAGE in proteomics and the wider investigation of protein:protein interactions with a focus on its use and potential in plant science. Major advantages and disadvantages of this technique when compared to other experimental strategies will be mentioned and a short overview about past and possible future applications of BN-PAGE in the plant sciences is provided.
The working principle of BN-PAGE
Electrophoretic separation
In the popular denaturing SDS-PAGE technique, the ionic detergent SDS functions both in solubilizing and denaturing the proteins as well as providing negative charge and therefore unidirectional mobility to the proteins during electrophoresis. In BN-PAGE these multiple functions are accomplished by different reagents. While denaturing is not desired, the solubilization of membrane protein complexes is necessary and is undertaken by mild non-ionic detergents like dodecylmaltoside, Triton X-100 or digitonin. Negative charges are attached to the solubilized protein complexes not by detergents but by the addition and binding of the negative charged dye Coomassie Blue G250. Dissociated dye, separated from the proteins during the gel run is substituted for by a high concentration of Coomassie Blue in the cathode buffer (Figure 1A).
Figure 1 The working principle of BN/SDS-PAGE. A) After solubilization, the mixture of different protein complexes is separated by BN-PAGE, according to their molecular weight. B) Following the gel run, the lane of the gel is excised and subjected to a denaturation solution (eg 1% SDS and 1% β-mercaptoethanol) so that the native protein complexes (underlayed in grey colour) dissociate to their constituent polypeptides. Loosely stuck in the pores of the gel, the subunits of the protein complexes (underlayed in red colour) remain at their position until they are forced electrophoretically into the second dimension gel. Due to the SDS used in the denaturation step (and residual Coomassie) the polypeptides are uniformly negatively charged and are separated according to their molecular weight in the gel. Subunits of a protein complex form a vertical row on the second dimension gel.
Preparing the sample
(a) Pre-fractionation and enrichment
The reason that BN-PAGE has been most commonly used for the analysis of the respiratory and photosynthetic protein complexes is likely the high abundance of them within the proteomes of these organelles. Relatively little sample material is required for a good visualization of these complexes as more than half of the total protein content of these organelles as estimated to be protein complexes. Low abundance complexes also exist in these organelles, but will remain in the background or may be masked in BN-PAGE separations by the dominant complexes. Successful use of the BN-technique for other applications may require pre-treatment to enrich the protein complex of interest. This could involve a simple concentration step (e.g. by dialysis or ultrafiltration), the removal of unwanted high abundant protein complexes (e.g. by affinity chromatography) or a fractionation step. Fractionation of the sample by a property other than size (e.g. by subcellular separations, ion-exchange chromatography, native IEF) can reduce complexity and therefore compensate for the limited range of sensitivity of BN-PAGE. Prior to the BN-PAGE, it might be necessary to change the buffer system in the sample depending on the method of pre-fractionation used.
(b) Solubilization of membrane samples
Compared to IEF, the collection of detergents tested for BN-PAGE is still relatively small. Finding suitable detergents for the solubilization of different protein complexes is a key for a wider application of BN-PAGE in the investigation of membrane protein complexes. A detergent suitable for a native solubilization of protein complexes prior to BN-PAGE must fulfil several prerequisites. It must be as mild as possible but able to disrupt lipid-lipid interactions without disturbing those between protein components in complexes. Disruption of certain lipid-protein associations is usually not desired because it may result in a weakening or even dissociation of protein complexes, especially if lipids are a structural component of the complex. The detergent should also not interfere with the electrophoresis process. Detergents used routinely for the solubilization of protein complexes include n-Dodecylmaltoside, Triton X100 and digitonin. Additionally, octylglucoside [2], Brij 96 [3] and saponin [3] have been used (Table 1). Other non-ionic detergents including Big CHAPS, C12E5/8, n-Decanoylsucrose and NP-40 may also be suitable for the native solubilization of membrane protein complexes. Some of these reagents have a defined composition and are synthesized to high degrees of homogeneity, while others such as digitonin are complex mixtures purified from natural sources. Comparative testing of different detergents in various detergent-to-protein ratios in the presence of different salts is necessary to determine the optimal detergent and the right conditions for solubilization [4-6]. A salt of low ionic strength like aminocaproic acid or potassium acetate can further support the solubilization process of membrane complexes.
Table 1 Detergents successfully employed for the solubilization of protein complexes prior to BN-PAGE.
Detergent Sample Detergent concentration Reference
Dodecylmaltoside Mammalian 1.1% [1]
Mitochondria
Plant Mitochondria 1.5% [103]
Chloroplasts 2.0% [75]
Yeast Peroxisomes 0.5% and 1.0% [81]
Digitonin Mammalian 4 g/g protein [104]
Mitochondria
Plant Mitochondria 5.0 g/g protein [6]
Chloroplasts 1.5 g/g protein [41]
Yeast Microsomes 1.0% [80]
Yeast Peroxisomes 1.0% [81]
Triton X-100 Mammalian 1.0, 1.4 and 2.4% [105]
Mitochondria
Plant Mitochondria 0.5% [6]
Yeast Peroxisomes 0.2% and 1.0% [81]
Whole cellular lysates of different human cell lines 0.1% [3]
Brij 96 Whole cellular lysates of different human cell lines 0.5% [3]
Octylglucoside Tobacco microsomes 40 mM [2]
Saponin Whole cellular lysates of different human cell lines 1.0% [3]
The analysis of mitochondrial respiratory chain complexes utilizing BN-PAGE has shown the influence of the detergent on the results and the conclusions drawn from these results. Classically, n-dodecylmaltoside and Triton X100 have been used exclusively to solubilize the respiratory chain complexes of the inner mitochondrial membrane for BN-PAGE. Using these detergents, only bands representing singular respiratory complexes I, II, III, IV and V can be found on the gels. It was assumed that the complexes found on the gels reflect the situation in vivo, supporting the "liquid state model" of individual respiratory protein complexes which are free to laterally diffuse in the inner mitochondrial membrane. However, since the introduction of digitonin for the solubilization of the respiratory protein complexes, the model of the respiratory chain has changed considerably. Stoichiometric association of several components of the electron transfer chain clearly indicate a "solid state model" for the respiratory chain of several organisms. In this model, the protein complexes are not randomly distributed within the inner mitochondrial membrane but form different types of defined arrangements which are called supercomplexes [5]. As the singular respiratory complexes solubilized by DDM and Triton X100 are active, the biological function of the supercomplexes remains unclear, but pictures from electron microscopy strongly support the solid state model [7].
The different membranes within eukaryotic cells each possesses a distinct lipid composition and lipid to protein ratio and different detergents in different concentrations will be needed for the optimal solubilization of their protein complexes. Additionally, plants are able to alter the lipid composition of their membranes to compensate for changes in the climatic conditions like cold stress [8,9]. This effect may also influence the suitability of detergents.
(c) Preparation of soluble complexes
Although theoretically less troublesome to investigate than membrane protein complexes, relatively few publications so far use BN-PAGE for the analysis of soluble protein complexes. Direct application of soluble plant cell extracts to BN-gels can result in visible smearing and/or excessive narrowing or widening of the lanes. This is often due to an inappropriate salt concentration for the running conditions of the gel. A buffer exchange to standard BN conditions, for example by dialysis, can help to reduce these symptoms. This, however, has to be undertaken cautiously as soluble protein complexes tend to be more sensitive to disruption by high salt concentrations than membrane complexes and even low salt concentrations may result in a dissociation of soluble proteins [10]. Stability of some protein complexes may also be increased if Coomassie Blue is not added to the sample prior to the gel run. Lengthy exposure in too much Coomassie dye can result in the dissociation of protein complexes. In these cases, the amount of dye in the cathode buffer is usually sufficient for the electrophoresis run [11].
Maximising the resolution of BN separation
To maximise resolution of the BN-gel, it usually consists of an acrylamide concentration gradient from 3–5% (w/v) at the top (cathodic end) to 13–16% (w/v) at the bottom (anodic end) (Figure 1A). Depending on the acrylamide gradients and their size, protein complexes appear to become lodged at different positions in the gel during the run, resulting in an apparent separation of the complexes according to molecular mass. BN-PAGE was initially invented for the investigation of respiratory chain components, thus it was optimised for the separation of 0.1 to 1 MDa protein complexes. With a reduction of the acrylamide concentration in the separating gel it is possible to resolve complexes of up to 3–4 MDa on BN-PAGE. The use of agarose to replace polyacrylamide in the separating gel can be considered if even larger complexes or supercomplexes (up to ~10 MDa) are expected [10].
Second and third dimensions for BN gels
In combination with an SDS gel system as second dimension, BN-PAGE is often used as the first dimension of multiple dimension gels (Figure 1B). For a second dimension, a lane of the BN-gel can be cut out and placed horizontally on the second dimension gel (that is at 90° to the direction of the first electrophoretic separation). To aid transfer of the protein complexes/proteins into the second dimension gel, the first dimension lane can be placed between the glass plates of the second dimension gel on its initial assembly. The gel is then cast with the lane already in position and it is later completely embedded in the stacking gel, thus ensuring a smooth transition of the proteins from the first into the second dimension. In this case it is useful to make the second dimension gel thinner (e.g. 1 mm) than the first dimension (e.g. 1.5 mm) to ensure that the BN gel strip stays in place during this procedure. A stacking gel is necessary for focussing of the now vertical bands of the BN gel lane. Depending on the type of second dimension used, the BN-gel lane can be treated prior to assembly of the second dimensions with different reagents (see below). BN-PAGE can also be combined with IEF/SDS-PAGE, resulting in a 3D gel system allowing very high resolution of protein samples (see below).
BN/SDS-PAGE
The BN gel strip should be incubated prior to the casting of the second dimension in a buffer containing SDS and 2-mercaptoethanol (Figure 1B). This step ensures complete denaturation of the protein complexes necessary for the subsequent separation of their subunits. The stacking gel is cast around the first dimension BN gel lane and preferably uses the same gel buffer. For visualisation and separation of samples containing small proteins of interest (<15 kDa) best results are obtained if the second dimension gel is a Tricine system consisting of a stacking, spacer and separating gel [12]. Good results can also be obtained using a standard Glycine based system [13] when analysis is concentrating on larger proteins (>15 kDa). A typical BN/SDS-PAGE from a plant mitochondrial membrane sample is shown in Figure 2A.
Figure 2 Second and third dimensions for BN-PAGE of Arabidopsis mitochondrial protein complexes. A) BN-PAGE of a digitonin-solubilized fraction followed by denaturing and reducing SDS-PAGE (BN/SDS-PAGE); B) BN-PAGE of Digitonin-solubilized protein complexes followed by a second BN-PAGE (BN/BN-PAGE) employing n-Dodecylmaltoside to introduce slightly less native conditions inducing partial dissociation of protein supercomplexes; C) IEF/SDS-PAGE of the ATP-Synthase band of a dodecylmaltoside-solubilized sample. The band has been cut out of the gel and the proteins have been subsequently electroeluted under denaturing conditions followed by a precipitation step to concentrate and purify the proteins for isoelectric focussing. The image (c) was reproduced from figure 1 in [15] with permission from Wiley-VCH.
BN/BN-PAGE
Different non-ionic detergents appear to alter the stability of protein complexes. For example, digitonin can be used for the solubilization of protein complexes and it has been shown to allow stabilization of supercomplexes in particular samples, while dodecylmaltoside destablizes some of these complex interactions [6,7]. Treatment of a first dimension BN-lane of digitonin-solubilized supercomplexes with DDM leads to the dissociation of these structures into single protein complexes. On the resulting BN/BN gels, intact supercomplexes and protein complexes form a diagonal line, whereas those which are specifically destabilized under the conditions of the second gel dimension migrate below the diagonal (Figure 2B). Alternatively to such detergent treatments, the BN strip of the first gel dimension can also be subjected to other mildly denaturing conditions like elevated temperature, salt concentration, reductants or other suitable chemicals. By comparison of the electrophoretic mobility of the decomposition products with that of the intact singular complexes on BN/BN gels, the protein complexes forming the larger supercomplexes can be identified. This can be confirmed by mass spectrometry of the complex components. If BN/BN gels are intended to be used to perform in-gel activity stains, it can be necessary to cast the second dimension gel first and, after complete polymerization, place the first dimension lane on top of it. This avoids contact of the complexes with highly reactive APS and TEMED and can help to maintain enzymatic activities [14].
BN/IEF/Tricine-SDS-PAGE
BN gels can also be used to reduce the complexity of a sample for subsequent IEF/SDS-PAGE. A band is cut from a BN gel and the proteins then electroeluted. A standard 2D-IEF/SDS-PAGE can be performed on this protein sample subsequently, separating only the subunits of the protein complex of choice [15] (Figure 2C). This so-called 3D gel system has the advantage of a higher resolution power over normal BN/SDS-PAGE as the polypeptide subunits are resolved by IEF and SDS-PAGE dimensions. It notably allows the separation of isoforms of subunits of the same molecular weight if they differ in their isoelectric points.
Staining BN gels
After thorough removal of the Coomassie dye in a solution of 40–50% methanol and 10% acetic acid, first dimension BN-gels can be stained with all commonly applied procedures like silver stains, classical and colloidal Coomassie stains and fluorescent dyes. Differential staining procedures using covalently bound fluorescent dyes of different colours (Differential gel electrophoresis, DIGE) have also been employed successfully on BN gels [16] (Figure 3A). It is, however, important to use fluorescent dyes that covalently bind to lysine residues rather than cysteine residues, as the latter may tend to destabilize disulphide linkages between and within protein complexes prior to BN separation which makes the sample unsuitable for analysis of native complex structure (H. Eubel, unpublished results).
Figure 3 Differential gel electrophoresis (DIGE) stains, immuno-stains and in-gel activity stains of protein complexes separated by BN-PAGE. A) DIGE staining of mitochondrial protein complexes on BN-SDS gels. Proteins displayed in red are from an Arabidopsis mutant with a reduced abundance of complex I, green proteins are subunits of the wild-type sample. If mutant and wild-type proteins are present in the same amount, they are coloured yellow. This image was reproduced from Figure 8a in 25 with permission from Elsevier. B) Immunoblot of two lanes of a BN-gel and a Coomassie-colloidal stained reference lane (CC). Lane 1 has been probed with an antibody directed against respiratory complex I, lane two with an antibody against complex III. This image was reproduced from Figure 2 in [103] with permission from Blackwell Publishing. C) In-gel activity stains of Arabidopsis respiratory complexes I, II and IV. An unstained (US) and Coomassie colloidal (CC) stained gel lane are given as references. This image was reproduced from Figure 2 in [35] with permission from Elsevier. D) Complex IV activity stained BN/BN gel of the same sample.
Western blotting of BN gels
First dimension BN gels can be successfully electroblotted onto nitrocellulose or PVDF membranes if a few requirements are met (Figure 3B). Firstly the gel run has to aborted at an early stage (i.e. after half to two third of the normal run time) to prevent the complexes from becoming stuck in the pores of the gel. Secondly, due to the high mobility of Coomassie and its high concentration in the gel, the cathode buffer should be exchanged for one containing no Coomassie after approximately two hours of the run to prevent excessive dye deposition on the membrane during transfer [11]. Alternatively, the membrane can be replaced a few times during the early stages of the blotting process to remove the bulk of the Coomassie dye. PVDF membranes generally have a lower affinity for Coomassie than nitrocellulose [11].
Identification of protein complexes
If a reference gel of the same sample type from a different species exists, protein complexes can often be identified by a simple comparison of the subunit clusters on BN/SDS gels. Within a kingdom, orthologous protein complexes often have similar amount of subunits with a similar distribution of molecular masses. This gives a good first hint of their identity. Because of evolutionary distance, a comparison between species of different kingdoms is often difficult. The most effective way to identify unknown protein complexes is mass spectrometry. Samples directly cut out of a first dimension BN gel or a BN/BN gel can be subjected to complex mixture tandem mass spectrometry to identify a series of proteins in the complex. Alternatively, individual subunits derived from a BN/SDS PAGE gel can be used for identification by tandem mass spectrometry or peptide mass fingerprinting. Often, the identification of a few subunits is sufficient for a reasonable identification of the complex. In a complementary strategy to consider complex function and also to aid identification, colorimetric enzyme assays can be performed as in-gel activity stains of first dimension BN [17,18] (Figure 3C) or second dimension BN/BN gels [14] (Figure 3D).
Known limitations of BN-PAGE
Detergents
For membrane protein complexes, the use of BN-PAGE is only limited by the availability of a suitable detergent for the native solubilization of the protein complexes. However, to date, only a very restricted collection of detergents has been tested for this application and, unfortunately, so far there seems to be no single detergent which is suitable for solubilization prior to BN-PAGE for all protein complexes of interest (Table 1). More extensive testing of detergents is required to determine the real limitations of the BN technique for the analysis of membrane protein complexes.
Resolution
The number of spots from a sample resolved on a 2D BN/SDS PAGE gel is normally smaller than from the same sample resolved by 2D IEF/SDS PAGE. This is mainly due to the fact that proteins are not distributed over the entire 2D area in BN/SDS PAGE, but arranged in vertical rows. Even more than in other gel based proteomic approaches, this dictates a reduction in sample complexity for BN separations. Single proteins or complexes of a molecular weight <100 kDa are not well resolved in BN-PAGE due to the high abundance of proteins in that size range and the limited separation distance resulting from the acrylamide gradient. Another resolution drawback seems to be a rather limited dynamic range, leading to a focus on the most prominent protein complexes and an under-representation of lower abundant complexes on BN gels.
Artefacts
Co-migration of proteins on BN-gels is not a final proof of native association as physically distinct complexes may have similar molecular masses and therefore may appear in the same protein bands. However, the resolution of this technique is much higher than conventional gel filtration chromatography approaches, and if the sample has been well fractionated to limit complexity we consider it an excellent line of evidence for association. Confirmation can be obtained using immunoprecipitations or co-migration under different conditions (eg non-sized-based chromatography). Weak or transient interactions between proteins or protein complexes constitute another challenge to the system. Even if an appropriate detergent is found, it inevitably will weaken the association of the proteins. The addition of Coomassie might lead to dissociation of fragile protein complexes, because negative charges on protein subunits of a protein complex can lead to electrical repulsion. The binding capacity of Coomassie and therefore the electrophoretic mobility depends on the physical and chemical properties of the proteins, e.g. size, shape, hydrophobicity, post-transcriptional modifications and isoelectric point. This can make it hard to judge the precise molecular mass of a band on a BN-gel [19].
Current biological applications of BN-PAGE
Studies of mitochondrial and chloroplast electron transport chains
BN-PAGE has been extensively used to investigate the structure of the respiratory chain in plants. In mammalian mitochondrial samples all components of the OXPHOS protein complexes could be resolved in the initial investigations using BN-PAGE [1]. However, in plant mitochondria only complexes I, III and V were present on the first BN-PAGE gels and dissociation of particular complexes was noted such as the release of the F1 component of the complex V [20]. The assembly of complex I of maize mitochondria [21] as well as Chlamydomonas has been studied by BN-PAGE [22]. Also, the composition of complex I and function of plant specific subunits in Arabidopsis have been investigated [23-25]. Resolution of complex III has allowed investigation of the core proteins of cytochrome reductase from potato [20] and the co-evolution of these proteins in different organisms [26]. Proteins responsible for the development of cytoplasmic male sterility (CMS) have been identified by BN-PAGE as being subunits of complex V [27,28]. BN-PAGE has also been employed to monitor the purity of mitochondria isolations from Brassica [29] and Arabidopsis [30,31]. Since the completion of the Arabidopsis and rice genome sequencing project and the onset of gel based proteomics, BN-PAGE has also been employed for more complete investigations of the respiratory chain [15,32], aiming to provide a broader overview of plant mitochondrial protein complexes and their subunit composition. However, complexes II and IV were still missing on most of these gels. This changed with the introduction of digitonin for the solubilization of mitochondrial membrane protein complexes. Digitonin not only led to the stabilisation of these other complexes on the gels and the discovery of plant specific subunits within them [6,33], but it also prevented the dissociation of respiratory supercomplexes, formed by components of the electron transfer chain [6,14,34,35].
Several studies have also used BN-PAGE for the analysis of photosynthetic protein complexes and the plastidic F0F1-ATPase [36-50]. Again, the utilization of digitonin has facilitated the solubilization of protein supercomplexes [41], confirming results obtained by crystallography and electron microscopy followed by single particle analysis (for a review see [51]). Other research utilizing this technique in plant organelle research includes the investigation of the NAD(P)H dehydrogenase complex of thylakoid and etioplast membranes [52-56].
Other applications in plants
In a number of publications BN-PAGE has also been used for the survey of the mitochondrial and plastidic protein import apparatus [15,57-62] and the thylakoid protein insertion systems [63-66]. It has been employed for the characterization of the tobacco plastid-encoded plastid RNA-polymerase complex [67,68] and plastidic omega-3 saturases [69] as well as a 350 kDa plastid ClpP protease complex [67]. BN-PAGE has also been used for the analysis of the cytochrome c maturation complex [70], an acetyl-coenzyme A carboxylase [71], an isovaleryl coenzyme A dehydrogenase [72], the formate dehydrogenase complex [73] and several other soluble mitochondrial protein complexes [74]. Protein complexes of the plasma membrane of spinach have been successfully resolved by BN-PAGE [75] as have complexes in the peribacteroid membrane from Lotus japanicus root nodules [76] and in the microsomes of Arabidopsis [77]. A ~200 kDa protein complex from peanut representing a putative food allergen has also been characterised by BN-PAGE [78].
Diverse applications in bacteria, yeast and mammals
Using several different detergents and antibodies, Camancho-Carvajal et al. [3] identified several protein complexes in whole cell lysates of human cell lines without previous subfractionation and enrichment. Applying BN-PAGE and other techniques, it was also found that the nucleosome assembly protein NAP-2 forms part of multiprotein complexes in human HeLa cell cultures [79]. Shibatani et al. [80] discovered that a mammalian oligasaccharyltransferase complex forms an integral component of the ER translocation machinery. More than ten protein complexes were discovered in the peroxisomal membrane of the yeast H. polymorpha [81]. In several publications, BN-PAGE has been used to investigate presenelin/γ-secretase complexes, associated with Alzheimer's desease [82-87]. Krall et al. resolved different VirB protein complexes of A. tumefaciens [88]. The E. coli Twin Arginine Translocase [89] has been analysed by BN-PAGE as well as the human glycine receptor, a member of the ligand-gated ion channel (LGIC) superfamily I, expressed in Xenopus oocytes [90]. Recently, a systematic analysis of the E. coli envelope revealed the presence of 43 protein complexes [91], thus making it the most comprehensive approach to membrane protein complex analysis employing BN-PAGE to date. As a result of its broad analysis, this study was also able to assign several proteins with unknown functions to specific protein complexes.
Future opportunities for BN in plant research
It is becoming more and more apparent that protein complexes are not an exception, but are a basic working principle of the cell. BN-PAGE can provide data not only on protein complex composition, it can also be used to increase our knowledge of the protein content of a proteome of choice. The use of BN-PAGE in the investigation of plant membrane complexes has already revealed numerous differences in the plant metabolism when compared to that of bacteria, fungi & mammals. Taking the lifestyle of plants into consideration, a large number of different or unique metabolic pathways might be expected to be present. These pathways almost certainly demand the existence of plant specific protein complexes. So far, most protein complexes characterised to date in plants are involved in energy metabolism and the genetic replication system. The large number of different pores, transporters and carriers present in the membranes of plant cells are still not characterized at a biochemical level, the same is true for the many protein complexes involved in signal perception and signal transduction. Due to its relatively lower resolution, when compared to other gel based techniques, the future of BN-PAGE will likely be linked to the investigation of defined subproteomes of the cell.
BN-PAGE to further explore plant proteomes
Approaches to separate and analyse proteomes will continue to be developed to complement IEF-SDS/PAGE. The combination of techniques will most probably result in a higher coverage of any given proteome. On one hand non-gel approaches are becoming a dominant approach to separate and analyse peptides from complex mixtures by mass spectrometry [92,93]. These have the advantage of better representing the hydrophobic proteome, but suffer from losses in the quantification of members of the proteome compared to IEF-SDS/PAGE. On the other hand, an approach like BN-PAGE can maintain quantification and greatly improve hydrophobic protein identification. A direct comparison of the protein sets of the Arabidopsis mitochondria proteome derived either by IEF/SDS-PAGE [30,94] or BN/SDS-PAGE [6,23,33] reveals only a slight difference in the average of the Grand Average of Hydropathicity (GRAVY) scores of the sets (-0.23 and -0.13). The root mean square, however, is two times higher for the BN derived-set (0.40) than for the IEF derived-set (0.20). This clearly indicates a much broader range in the GRAVY scores for the set derived from BN/SDS-PAGE. In the BN set, 22% of proteins possess GRAVY scores above zero, whereas this is only the case for 7% of proteins identified by IEF.
It might be argued that the technical aspects of BN-PAGE make it unlikely to be widely adopted outside specialist laboratories. To date, users of BN-PAGE still have to prepare their own first dimension gels. Being an acrylamide gradient, it can be compared to preparing IEF tube gels in terms of labour and time requirements. However, in contrast to IEF, there is no special equipment needed to perform BN-PAGE. Any common vertical gel system can easily be modified to suit the requirements for BN-PAGE. The original protocol for IEF/SDS-PAGE [95,96] used self cast, labour intensive tube gels for the first dimension. Since the introduction of commercially available IEF systems employing immobilized pH-gradient (IPG) strips, the use of tube gels has greatly reduced. Handling of the IPG strips is easy and, because of mass-production, they ensure a higher reproducibility. Several kinds of IPG strips are commercially available, giving the researcher the opportunity to choose the appropriate pH range and gradient for his application. More commercial products for BN could be produced if the market expanded.
BN-PAGE for the analysis of protein:protein interactions in plants
Most proteome analyses in plants to date either considers the location of proteins within cells or are a differential analysis of expression in response to development, environment or disease. This lacks the information about the interaction pattern of these proteins, which is a very important aspect of understanding the processes carried out by the members of this proteome. Popular techniques to determine the interaction partners of proteins are the yeast-two-hybrid technique (Y2H), affinity chromatography (eg tandem affinity purification [TAP]), fluorescence resonance energy transfer (FRET), bioluminescence resonance energy transfer (BRET), bimolecular fluorescent complementation, high-throughput mass spectrometric protein complex identification (HMS-PCI), size exclusion chromatography (SEC) and cross-linking experiments. Most of these techniques require the organism of choice to be accessible to genetic engineering. With the exception of SEC and some cross-linking techniques, all these approaches have in common that they focus on a single protein and determine its binding partners.
The yeast two-hybrid system has been used extensively to study protein:protein interactions in S. cerevisae [97,98] but is known to produce a relatively high number of false positive results [98]. In one of the largest uses of HMS-PCI, 493 bait proteins in S. cerevisae were found to be involved in 3617 interactions after correction for false positives, 74% of these interactions could be confirmed by immuno-precipitation experiments [99]. These numbers give insight into the complexity of protein networks within a cell and emphasize the need for complementary ways for the analysis of in vivo protein interactions. Being a preparative technique, SEC has a lower resolution compared to BN-PAGE. Also, the micelle size of the detergent and possible unspecific aggregations have an influence on the apparent size of a protein complex in SEC. In BN-PAGE, these effects are reduced by the presence of Coomassie blue and aminocaproic acid [2]. Chemical crosslinkers are usually only suitable for the analysis of small protein complexes and cross-linking experiments often suffer from poor yields of the interacting proteins [100]. This often necessitates long incubation times that in turn increase the probability of the generation of artefacts or leading to a destabilization of the complex. With the introduction of photo-inducible cross-linking of proteins, the risk of artefact-formation has been reduced considerably [101]. The introduction of cleavable cross-linkers allows the direct identification of protein interaction sites by mass spectrometry [102]. However, to our knowledge, this technique is still limited to relatively small protein complexes. It also allows no rating of the binding strength between the proteins within the complex due to the covalent modifications introduced by the technique. And of course, great care has to be taken in the choice of the cross-linking reagent to avoid false positive results. BN offers a broader approach as the composition of many small and large protein complexes and therefore the interaction patterns of all their protein-subunits in a relatively complex sample can be determined in a single experiment. Further, protein overexpression and/or chemical modification is not involved, which often allows high yields of native complexes for analysis.
Conclusion
Since its invention, BN-PAGE has had an enormous impact on the investigation of the respiratory chains and photosynthetic complexes in a range of organisms. When used in a 2D system, BN-PAGE allows an assignment of proteins to their protein complexes and display of highly hydrophobic proteins in two dimensions. Recently, BN-PAGE is beginning to be applied to different types of samples and with many different aims. These range from the assessment of the oligomeric state of protein complexes to the analysis of complex mixtures of protein complexes. BN-PAGE, together with other techniques, might be the ideal tool to begin study of new protein complexes in plants and gain a deeper understanding of the unique aspects of protein-protein interactions in plant cellular processes.
Competing interests
The author(s) declare that they have no competing interests.
Acknowledgements
AHM is funded as an Australian Research Council (ARC) QEII Fellow and thanks ARC for funding under the ARC Centre of Excellence Program. HPB is supported by the Deutsche Forschungsgemeinschaft (grant Br1829-7/1).
==== Refs
Schagger H Vonjagow G Blue Native Electrophoresis for Isolation of Membrane-Protein Complexes in Enzymatically Active Form Analytical Biochemistry 1991 199 223 231 1812789 10.1016/0003-2697(91)90094-A
Rivas S Romeis T Jones JDG The cf-9 disease resistance protein is present in an similar to 420-kilodalton heteromultimeric membrane-associated complex at one molecule per complex Plant Cell 2002 14 689 702 11910014 10.1105/tpc.010357
Camacho-Carvajal MM Wollscheid B Aebersold R Steimle V Schamel WWA Two-dimensional blue native/SDS gel electrophoresis of multi-protein complexes from whole cellular lysates - A proteomics approach Molecular & Cellular Proteomics 2004 3 176 182 14665681 10.1074/mcp.T300010-MCP200
Arnold I Pfeiffer K Neupert W Stuart RA Schagger H Yeast mitochondrial F1F0-ATP synthase exists as a dimer: identification of three dimer-specific subunits EMBO J 1998 17 7170 7178 9857174 10.1093/emboj/17.24.7170
Schagger H Respiratory chain supercomplexes of mitochondria and bacteria Biochim Biophys Acta 2002 1555 154 159 12206908
Eubel H Jansch L Braun HP New insights into the respiratory chain of plant mitochondria. Supercomplexes and a unique composition of complex II Plant Physiology 2003 133 274 286 12970493 10.1104/pp.103.024620
Dudkina NV Eubel H Keegstra W Boekema EJ Braun HP Structure of a mitochondrial supercomplex formed by respiratory-chain complexes I and III P Natl Acad Sci USA P Natl Acad Sci USA 2005 102 3225 3229 10.1073/pnas.0408870102
Uemura M Joseph RA Steponkus PL Cold Acclimation of Arabidopsis thaliana (Effect on Plasma Membrane Lipid Composition and Freeze-Induced Lesions) Plant Physiol 1995 109 15 30 12228580
Uemura M Steponkus PL Effect of Cold Acclimation on the Lipid Composition of the Inner and Outer Membrane of the Chloroplast Envelope Isolated from Rye Leaves Plant Physiol 1997 114 1493 1500 12223783
Henderson NS Nijtmans LGJ Lindsay JG Lamantea E Zeviani M Holt LJ Separation of intact pyruvate dehydrogenase complex using blue native agarose gel electrophoresis Electrophoresis 2000 21 2925 2931 11001305 10.1002/1522-2683(20000801)21:14<2925::AID-ELPS2925>3.0.CO;2-2
Schagger H Blue-native gels to isolate protein complexes from mitochondria Methods in Cell Biology, Vol 65 2001 65 231 244
Schagger H Vonjagow G Tricine Sodium Dodecyl-Sulfate Polyacrylamide-Gel Electrophoresis for the Separation of Proteins in the Range from 1-Kda to 100-Kda Analytical Biochemistry 1987 166 368 379 2449095 10.1016/0003-2697(87)90587-2
Laemmli UK Cleavage of structural proteins during the assembly of the head of bacteriophage T4 Nature 1970 227 680 685 5432063 10.1038/227680a0
Eubel H Heinemeyer J Braun HP Identification and characterization of respirasomes in potato mitochondria Plant Physiology 2004 134 1450 1459 15064371 10.1104/pp.103.038018
Werhahn W Braun HP Biochemical dissection of the mitochondrial proteome from Arabidopsis thaliana by three-dimensional gel electrophoresis Electrophoresis 2002 23 640 646 11870776 10.1002/1522-2683(200202)23:4<640::AID-ELPS640>3.0.CO;2-F
Perales M Parisi G Fornasari M Colaneri A Villarreal F Gonzalez-Schain N Echave J Gomez-Casati D Braun HP Araya A Zabaleta E Gamma carbonic anhydrase like complex interact with plant mitochondrial complex I Plant Molecular Biology 2004 56 947 957 15821992 10.1007/s11103-004-6324-z
Zerbetto E Vergani L DabbeniSala F Quantification of muscle mitochondrial oxidative phosphorylation enzymes via histochemical staining of blue native polyacrylamide gels Electrophoresis 1997 18 2059 2064 9420170 10.1002/elps.1150181131
Jung C Higgins CM Xu Z Measuring the quantity and activity of mitochondrial electron transport chain complexes in tissues of central nervous system using blue native polyacrylamide gel electrophoresis Anal Biochem 2000 286 214 223 11067743 10.1006/abio.2000.4813
Schamel WWA Biotinylation of protein complexes may lead to aggregation as well as to loss of subunits as revealed by Blue Native PAGE Journal of Immunological Methods 2001 252 171 174 11334977 10.1016/S0022-1759(01)00357-X
Jansch L Kruft V Schmitz UK Braun HP Cytochrome-C Reductase from Potato Does Not Comprise 3 Core Proteins but Contains an Additional Low-Molecular-Mass Subunit Eur J Biochem Eur J Biochem 1995 228 878 885
Karpova OV Newton KJ A partially assembled complex I in NAD4-deficient mitochondria of maize Plant Journal 1999 17 511 521 10.1046/j.1365-313X.1999.00401.x
Cardol P Matagne RF Remacle C Impact of mutations affecting ND mitochondria-encoded Subunits on the activity and assembly of complex I in chlamydomonas. Implication for the structural organization of the enzyme J Mol Biol J Mol Biol 2002 319 1211 1221
Heazlewood JL Howell KA Millar AH Mitochondrial complex I from Arabidopsis and rice: orthologs of mammalian and fungal components coupled with plant-specific subunits Bba-Bioenergetics Bba-Bioenergetics 2003 1604 159 169
Millar AH Mittova V Kiddle G Heazlewood JL Bartoli CG Theodoulou FL Foyer CH Control of ascorbate synthesis by respiration and its implications for stress responses Plant Physiol 2003 133 443 447 14555771 10.1104/pp.103.028399
Perales M Eubel H Heinemeyer J Colaneri A Zabaleta E Braun HP Disruption of a nuclear gene encoding a mitochondrial gamma carbonic anhydrase reduces complex I and supercomplex I + III2 levels and alters mitochondrial physiology in Arabidopsis J Mol Biol 2005 350 263 277 15935378 10.1016/j.jmb.2005.04.062
Brumme S Kruft V Schmitz UK Braun HP New insights into the co-evolution of cytochrome c reductase and the mitochondrial processing peptidase J Biol Chem J Biol Chem 1998 273 13143 13149 10.1074/jbc.273.21.13143
Heazlewood JL Whelan J Millar AH The products of the mitochondrial orf25 and orfB genes are F-o components in the plant F1Fo ATP synthase Febs Letters 2003 540 201 205 12681508 10.1016/S0014-5793(03)00264-3
Sabar M Gagliardi D Balk J Leaver CJ ORFB is a subunit of F1FO-ATP synthase: insight into the basis of cytoplasmic male sterility in sunflower Embo Reports 2003 4 381 386 12671689 10.1038/sj.embor.embor800
Mihr C Baumgartner M Dieterich JH Schmitz UK Braun HP Proteomic approach for investigation of cytoplasmic male sterility (CMS) in Brassica Journal of Plant Physiology 2001 158 787 794 10.1078/0176-1617-00292
Kruft V Eubel H Jansch L Werhahn W Braun HP Proteomic approach to identify novel mitochondrial proteins in Arabidopsis Plant Physiology 2001 127 1694 1710 11743114 10.1104/pp.127.4.1694
Hausmann N Werhahn W Huchzermeyer B Braun HP Papenbrock J How to document the purity of mitochondria prepared from green tissue of pea, tobacco and Arabidopsis thaliana Phyton-Annales Rei Botanicae 2003 43 215 229
Heazlewood JL Howell KA Whelan J Millar AH Towards an analysis of the rice mitochondrial proteome Plant Physiology 2003 132 230 242 12746528 10.1104/pp.102.018986
Millar AH Eubel H Jansch L Kruft V Heazlewood JL Braun HP Mitochondrial cytochrome c oxidase and succinate dehydrogenase complexes contain plant specific subunits Plant Molecular Biology 2004 56 77 90 15604729 10.1007/s11103-004-2316-2
Krause F Reifschneider NH Vocke D Seelert H Rexroth S Dencher NA "Respirasome"-like supercomplexes in green leaf mitochondria of spinach J Biol Chem J Biol Chem 2004 279 48369 48375 10.1074/jbc.M406085200
Eubel H Heinemeyer J Sunderhaus S Braun HP Respiratory chain supercomplexes in plant mitochondria Plant Physiology and Biochemistry 2004 42 937 942 15707832 10.1016/j.plaphy.2004.09.010
Zhang HM Whitelegge JP Cramer WA Ferredoxin : NADP(+) oxidoreductase is a subunit of the chloroplast cytochrome b(6)f complex J Biol Chem J Biol Chem 2001 276 38159 38165
Kugler M Jansch L Kruft V Schmitz UK Braun HP Analysis of the chloroplast protein complexes by blue-native polyacrylamide gel electrophoresis (BN-PAGE) Photosynthesis Research 1997 53 35 44 10.1023/A:1005882406718
Kugler M Kruft V Schmitz UK Braun HP Characterization of the PetM subunit of the b(6)f complex from higher plants Journal of Plant Physiology 1998 153 581 586
Huang D Everly RM Cheng RH Heymann JB Schagger H Sled V Ohnishi T Baker TS Cramer WA Characterization of the Chloroplast Cytochrome-B(6)F Complex as a Structural and Functional Dimer Biochemistry 1994 33 4401 4409 8155658 10.1021/bi00180a038
Seelert H Dencher NA Muller DJ Fourteen protomers compose the oligomer III of the proton-rotor in spinach chloroplast ATP synthase J Mol Biol J Mol Biol 2003 333 337 344
Heinemeyer J Eubel H Wehmhoner D Jansch L Braun HP Proteomic approach to characterize the supramolecular organization of photosystems in higher plants Phytochemistry 2004 65 1683 1692 15276430 10.1016/j.phytochem.2004.04.022
Hou CX Dirk LMA Williams MA Inhibition of peptide deformylase in Niconana tabacum leads to decreased D1 protein accumulation ultimately resulting in a reduction of photosystem II complexes Am J Bot Am J Bot 2004 91 1304 1311
Aro EM Suorsa M Rokka A Allahverdiyeva Y Paakkarinen V Saleem A Battchikova N Rintamaki E Dynamics of photosystem II: a proteomic approach to thylakoid protein complexes J Exp Bot J Exp Bot 2005 56 347 356 10.1093/jxb/eri041
Ciambella C Roepstorff P Aro EM Zolla L A proteomic approach for investigation of photosynthetic apparatus in plants Proteomics 2005 5 746 757 15682463 10.1002/pmic.200401129
Poetsch A Neff D Seelert H Schagger H Dencher NA Dye removal, catalytic activity and 2D crystallization of chloroplast H+-ATP synthase purified by blue native electrophoresis Bba-Biomembranes Bba-Biomembranes 2000 1466 339 349
Neff D Dencher NA Purification of multisubunit membrane protein complexes: Isolation of chloroplast FoF1-ATP synthase, CFo and CF1 by blue native electrophoresis Biochem Bioph Res Co Biochem Bioph Res Co 1999 259 569 575 10.1006/bbrc.1999.0820
Li BB Guo JK Zhou Y Zhang ZZ Zhang LX Blue native gel electrophoresis analysis of chloroplast pigment protein complexes Progress in Biochemistry and Biophysics 2003 30 639 643
Plucken H Muller B Grohmann D Westhoff P Eichacker LA The HCF136 protein is essential for assembly of the photosystem II reaction center in Arabidopsis thaliana Febs Letters 2002 532 85 90 12459468 10.1016/S0014-5793(02)03634-7
Thidholm E Lindstrom V Tissier C Robinson C Schroder WP Funk C Novel approach reveals localisation and assembly pathway of the PsbS and PsbW proteins into the photosystem II dimer Febs Letters 2002 513 217 222 11904154 10.1016/S0014-5793(02)02314-1
Rexroth S Tittingdorf RMWMZ Schwassmann HJ Krause F Seelert H Dencher NA Dimeric H+-ATP synthase in the chloroplast of Chlamydomonas reinhardtii Bba-Bioenergetics Bba-Bioenergetics 2004 1658 202 211
Dekker JP Boekema EJ Supramolecular organization of thylakoid membrane proteins in green plants Biochimica et Biophysica Acta (BBA) - Bioenergetics 2005 1706 12 39 10.1016/j.bbabio.2004.09.009
Quiles MJ Garcia A Cuello J Separation by blue-native PAGE and identification of the whole NAD(P)H dehydrogenase complex from barley stroma thylakoids Plant Physiology and Biochemistry 2000 38 225 232 10.1016/S0981-9428(00)00740-3
Quiles MJ Garcia A Cuello J Comparison of the thylakoidal NAD(P)H dehydrogenase complex and the mitochondrial complex I separated from barley leaves by blue-native PAGE Plant Science 2003 164 541 547 10.1016/S0168-9452(03)00003-7
Quiles MJ Lopez NI Photoinhibition of photosystems I and II induced by exposure to high light intensity during oat plant growth - Effects on the chloroplast NADH dehydrogenase complex Plant Science 2004 166 815 823 10.1016/j.plantsci.2003.11.025
Rumeau D Becuwe-Linka N Beyly A Louwagie M Garin J Peltier G New subunits NDH-M, -N, and -O, encoded by nuclear genes, are essential for plastid Ndh complex functioning in higher plants Plant Cell 2005 17 219 232 15608332 10.1105/tpc.104.028282
Guera A de Nova PG Sabater B Identification of the Ndh (NAD(P)H-plastoquinone-oxidoreductase) complex in etioplast membranes of barley: Changes during photomorphogenesis of chloroplasts Plant and Cell Physiology 2000 41 49 59 10750708
Jansch L Kruft V Schmitz UK Braun HP Unique composition of the preprotein translocase of the outer mitochondrial membrane from plants J Biol Chem J Biol Chem 1998 273 17251 17257 10.1074/jbc.273.27.17251
Caliebe A Grimm R Kaiser G Lubeck J Soll J Heins L The chloroplastic protein import machinery contains a Rieske-type iron-sulfur cluster and a mononuclear iron-binding protein Embo Journal 1997 16 7342 7350 9405363 10.1093/emboj/16.24.7342
Kuchler M Decker S Hormann F Soll J Heins L Protein import into chloroplasts involves redox-regulated proteins Embo Journal 2002 21 6136 6145 12426385 10.1093/emboj/cdf621
Kikuchi S Hirohashi T Nakai M Analysis of Toc complex on chloroplast outer envelope membrane by Blue Native PAGE Plant and Cell Physiology 2003 44 S203 S203
Werhahn W Niemeyer A Jansch L Kruft V Schmitz UK Braun HP Purification and characterization of the preprotein translocase of the outer mitochondrial membrane from Arabidopsis. Identification of multiple forms of TOM20 Plant Physiology 2001 125 943 954 11161051 10.1104/pp.125.2.943
Werhahn W Jansch L Braun HP Identification of novel subunits of the TOM complex from Arabidopsis thaliana Plant Physiology and Biochemistry 2003 41 407 416 10.1016/S0981-9428(03)00047-0
Berghofer J Klosgen RB Two distinct translocation intermediates can be distinguished during protein transport by the TAT (Delta ph) pathway across the thylakoid membrane Febs Letters 1999 460 328 332 10544258 10.1016/S0014-5793(99)01365-4
Cline K Mori H Thylakoid Delta pH-dependent precursor proteins bind to a cpTatC-Hcf106 complex before Tha4-dependent transport Journal of Cell Biology 2001 154 719 729 11502764 10.1083/jcb.200105149
Molik S Karnauchov I Weidlich CE Herrmann RG Klosgen RB The Rieske Fe/S protein of the cytochrome b(6)/f complex in chloroplasts - Missing link in the evolution of protein transport pathways in chloroplasts? J Biol Chem J Biol Chem 2001 276 42761 42766 10.1074/jbc.M106690200
Klostermann E Helling ID Carde JP Schunemann D The thylakoid membrane protein ALB3 associates with the cpSecY-translocase in Arabidopsis thaliana Biochemical Journal 2002 368 777 781 12217076 10.1042/BJ20021291
Peltier JB Ytterberg J Liberles DA Roepstorff P van Wijk KJ Identification of a 350-kDa ClpP protease complex with 10 different Clp isoforms in chloroplasts of Arabidopsis thaliana J Biol Chem J Biol Chem 2001 276 16318 16327 10.1074/jbc.M010503200
Suzuki JY Ytterberg AJ Beardslee TA Allison LA Wijk KJ Maliga P Affinity purification of the tobacco plastid RNA polymerase and in vitro reconstitution of the holoenzyme Plant Journal 2004 40 164 172 15361150 10.1111/j.1365-313X.2004.02195.x
Sakamoto H Matsuda O Hashimoto T Iba K Blue native PAGE analysis of oligomeric state of chloroplast-localized Arabidopsis omega-3 desaturases Plant and Cell Physiology 2004 45 S116 S116
Giege P Rayapuram N Meyer EH Grienenberger JM Bonnard G CcmFc involved in cytochrome c maturation is present in a large sized complex in wheat mitochondria Febs Letters 2004 563 165 169 15063743 10.1016/S0014-5793(04)00291-1
Focke M Gieringer E Schwan S Jansch L Binder S Braun HP Fatty acid biosynthesis in mitochondria of grasses: Malonyl-coenzyme A is generated by a mitochondrial-localized acetyl-coenzyme A carboxylase Plant Physiology 2003 133 875 884 12972648 10.1104/pp.103.027375
Daschner K Couee I Binder S The mitochondrial isovaleryl-coenzyme A dehydrogenase of Arabidopsis oxidizes intermediates of leucine and valine catabolism Plant Physiology 2001 126 601 612 11402190 10.1104/pp.126.2.601
Bykova NV Stensballe A Egsgaard H Jensen ON Moller IM Phosphorylation of formate dehydrogenase in potato tuber mitochondria J Biol Chem J Biol Chem 2003 278 26021 26030 10.1074/jbc.M300245200
Giege P Sweetlove LJ Leaver CJ Identification of mitochondrial protein complexes in Arabidopsis using two-dimensional blue-native polyacrylamide gel electrophoresis Plant Molecular Biology Reporter 2003 21 133 144
Kjell J Rasmusson AG Larsson H Widell S Protein complexes of the plant plasma membrane resolved by Blue Native PAGE Physiologia Plantarum 2004 121 546 555 10.1111/j.1399-3054.2004.00354.x
Wienkoop S Saalbach G Proteome analysis. Novel proteins identified at the peribacteroid membrane from Lotus japonicus root nodules Plant Physiology 2003 131 1080 1090 12644660 10.1104/pp.102.015362
Drykova D Cenklova V Sulimenko V Volc J Draber P Binarova P Plant gamma-tubulin interacts with alpha beta-tubulin dimers and forms membrane-associated complexes Plant Cell 2003 15 465 480 12566585 10.1105/tpc.007005
Boldt A Fortunato D Conti A Petersen A Ballmer-Weber B Lepp U Reese G Becker WM Analysis of the composition of an immunoglobulin E reactive high molecular weight protein complex of peanut extract containing Ara h 1 and Ara h 3/4 Proteomics 2005 5 675 686 15714473 10.1002/pmic.200401150
Rodriguez P Ruiz MT Price GB Zannis-Hadjopoulos M NAP-2 is part of multi-protein complexes in HeLa cells J Cell Biochem J Cell Biochem 2004 93 398 408 10.1002/jcb.20163
Shibatani T David LL McCormack AL Frueh K Skach WR Proteomic analysis of mammalian oligosaccharyltransferase reveals multiple subcomplexes that contain sec61, TRAP, and two potential new subunits Biochemistry 2005 44 5982 5992 15835887 10.1021/bi047328f
Salomons FA Lutz MV Meisinger C Pfanner N Veehuis M van der Klei IJ Sorting of matrix proteins to peroxisomes in the methylotrophic yeast Hansenula polymorpha 2001
Culvenor JG Reeve SB Smith AI Masters CL Analysis of presenilin complex by blue native electrophoresis and mass spectrometry reveals linkage to cytoskeletal elements Neurobiology of Aging 2004 25 S553 S553 10.1016/S0197-4580(04)81831-1
Culvenor JG Ilaya NT Ryan MT Canterford L Hoke DE Williamson NA McLean CA Masters CL Evin G Characterization of presenilin complexes from mouse and human brain using Blue Native gel electrophoresis reveals high expression in embryonic brain and minimal change in complex mobility with pathogenic presenilin mutations Eur J Biochem Eur J Biochem 2004 271 375 385
Evin G Canterford LD Hoke DE Sharples RA Culvenor JG Masters CL Transition-state analogue gamma-secretase inhibitors stabilize a 900 kDa presenilin/nicastrin complex Biochemistry 2005 44 4332 4341 15766262 10.1021/bi0481702
Ilaya NT Evin G Masters CL Culvenor JG Nicastrin expression in mouse peripheral tissues is not co-ordinated with presenilin and is high in muscle J Neurochem J Neurochem 2004 91 230 237
Nyabi O Bentahir M Horre K Herreman A Gottardi-Littell N Van Broeckhoven C Merchiers P Spittaels K Annaert W De Strooper B Presenilins mutated at Asp-257 or Asp-385 restore Pen-2 expression and nicastrin glycosylation but remain catalytically inactive in the absence of wild type presenilin J Biol Chem J Biol Chem 2003 278 43430 43436 10.1074/jbc.M306957200
LaVoie MJ Fraering PC Ostaszewski BL Ye WJ Kimberly WT Wolfe MS Selkoe DJ Assembly of the gamma-secretase complex involves early formation of an intermediate subcomplex of Aph-1 and nicastrin J Biol Chem J Biol Chem 2003 278 37213 37222 10.1074/jbc.M303941200
Krall L Wiedemann U Unsin G Weiss S Domke N Baron C Detergent extraction identifies different VirB protein subassemblies of the type IV secretion machinery in the membranes of Agrobacterium tumefaciens P Natl Acad Sci USA P Natl Acad Sci USA 2002 99 11405 11410 10.1073/pnas.172390699
Barrett CML Mangels D Robinson C Mutations in subunits of the Escherichia coli twin-arginine translocase block function via differing effects on translocation activity or Tat complex structure J Mol Biol J Mol Biol 2005 347 453 463
Griffon N Buttner C Nicke A Kuhse J Schmalzing G Betz H Molecular determinants of glycine receptor subunit assembly Embo Journal 1999 18 4711 4721 10469650 10.1093/emboj/18.17.4711
Stenberg F Chovanec P Maslen SL Robinson CV Ilag L von Heijne G Daley DO Protein complexes of the Escherichia coli cell envelope J Biol Chem 2005 16079137
Link AJ Eng J Schieltz DM Carmack E Mize GJ Morris DR Garvik BM Yates JR Direct analysis of protein complexes using mass spectrometry Nature Biotechnology 1999 17 676 682 10404161 10.1038/10890
Graumann J Dunipace LA Seol JH McDonald WH Yates JRIII Wold BJ Deshaies RJ Applicability of Tandem Affinity Purification MudPIT to Pathway Proteomics in Yeast Mol Cell Proteomics 2004 3 226 237 14660704 10.1074/mcp.M300099-MCP200
Millar AH Sweetlove LJ Giege P Leaver CJ Analysis of the Arabidopsis mitochondrial proteome Plant Physiol 2001 127 1711 1727 11743115 10.1104/pp.127.4.1711
O'Farrell PH High resolution two-dimensional electrophoresis of proteins J Biol Chem 1975 250 4007 4021 236308
Klose J Protein mapping by combined isoelectric focusing and electrophoresis of mouse tissues. A novel approach to testing for induced point mutations in mammals Humangenetik 1975 26 231 243 1093965
Uetz P Giot L Cagney G Mansfield TA Judson RS Knight JR Lockshon D Narayan V Srinivasan M Pochart P Qureshi-Emili A Li Y Godwin B Conover D Kalbfleisch T Vijayadamodar G Yang MJ Johnston M Fields S Rothberg JM A comprehensive analysis of protein-protein interactions in Saccharomyces cerevisiae Nature 2000 403 623 627 10688190 10.1038/35001009
Ito T Chiba T Ozawa R Yoshida M Hattori M Sakaki Y A comprehensive two-hybrid analysis to explore the yeast protein interactome P Natl Acad Sci USA P Natl Acad Sci USA 2001 98 4569 4574 10.1073/pnas.061034498
Ho Y Gruhler A Heilbut A Bader GD Moore L Adams SL Millar A Taylor P Bennett K Boutilier K Yang LY Wolting C Donaldson I Schandorff S Shewnarane J Vo M Taggart J Goudreault M Muskat B Alfarano C Dewar D Lin Z Michalickova K Willems AR Sassi H Nielsen PA Rasmussen KJ Andersen JR Johansen LE Hansen LH Jespersen H Podtelejnikov A Nielsen E Crawford J Poulsen V Sorensen BD Matthiesen J Hendrickson RC Gleeson F Pawson T Moran MF Durocher D Mann M Hogue CWV Figeys D Tyers M Systematic identification of protein complexes in Saccharomyces cerevisiae by mass spectrometry Nature 2002 415 180 183 11805837 10.1038/415180a
Herrmann JM Westermann B Neupert W Analysis of protein-protein interactions in mitochondria by coimmunoprecipitation and chemical cross-linking Methods Cell Biol 2001 65 217 230 11381595
Fancy DA Kodadek T Chemistry for the analysis of protein-protein interactions: rapid and efficient cross-linking triggered by long wavelength light Proc Natl Acad Sci U S A 1999 96 6020 6024 10339534 10.1073/pnas.96.11.6020
Tang X Munske GR Siems WF Bruce JE Mass spectrometry identifiable cross-linking strategy for studying protein-protein interactions Anal Chem 2005 77 311 318 15623310 10.1021/ac0488762
Jansch L Kruft V Schmitz UK Braun HP New insights into the composition, molecular mass and stoichiometry of the protein complexes of plant mitochondria. Plant Journal 1996 9 357 368 8919912 10.1046/j.1365-313X.1996.09030357.x
Schagger H Pfeiffer K The ratio of oxidative phosphorylation complexes I-V in bovine heart mitochondria and the composition of respiratory chain supercomplexes The Journal Of Biological Chemistry 2001 276 37861 37867 11483615
Schagger H Pfeiffer K Supercomplexes in the respiratory chains of yeast and mammalian mitochondria Embo Journal 2000 19 1777 1783 10775262 10.1093/emboj/19.8.1777
|
16287510
|
PMC1308860
|
CC BY
|
2021-01-04 16:37:28
|
no
|
Plant Methods. 2005 Nov 16; 1:11
|
utf-8
|
Plant Methods
| 2,005 |
10.1186/1746-4811-1-11
|
oa_comm
|
==== Front
Reprod HealthReproductive Health1742-4755BioMed Central London 1742-4755-2-101628864710.1186/1742-4755-2-10ResearchBrazilian obstetrician-gynecologists and abortion: a survey of knowledge, opinions and practices Goldman Lisa A [email protected]ía Sandra G [email protected]íaz Juan [email protected] Eileen A [email protected] University of California, Berkeley, Department of Epidemiology, 140 Warren Hall #7360, Berkeley, CA 94720, USA2 Population Council, Regional Office for Latin America and the Caribbean, Panzacola #62-102, Col. Villa Coyoacán, Mexico City 04000, Mexico3 Population Council, Brazil Office, Caixa Postal 6509, Cidade Universitária, Campinas, São Paulo, Brazil, Cep. 13084-970, Brazil2005 15 11 2005 2 10 10 9 5 2005 15 11 2005 Copyright © 2005 Goldman et al; licensee BioMed Central Ltd.2005Goldman et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Abortion laws are extremely restrictive in Brazil. The knowledge, opinions of abortion laws, and abortion practices of obstetrician-gynecologists can have a significant impact on women's access to safe abortion.
Methods
We conducted a mail-in survey with a 10% random sample of obstetrician-gynecologists affiliated with the Brazilian Federation of Obstetricians and Gynecologists. We documented participants' experiences performing abortion under a range of legal and illegal circumstances, and asked about which abortion techniques they had experience with. We used chi-square tests and crude logistic regression models to determine which sociodemographic, knowledge-related, or practice-related variables were associated with physician opinion.
Results
Of the 1,500 questionnaires that we mailed out, we received responses from 572 (38%). Less than half (48%) of the respondents reported accurate knowledge about abortion law and 77% thought that the law should be more liberal. One-third of respondents reported having previous experience performing an abortion, and very few of these physicians reported having experience with manual vacuum aspiration (MVA) or with misoprostol with either mifepristone or methotrexate. Physicians that favored liberalization of the law were more likely to have correct knowledge about abortion law, and to be in favor of public funding for abortion services.
Conclusion
Brazilian obstetrician-gynecologists need more information on abortion laws and on safe, effective abortion procedures.
==== Body
Background
In Brazil, as in most of the Latin America and Caribbean (LAC) region, abortion is highly legally restricted. The Brazilian Penal Code dating back to 1940 states that abortion is illegal except when performed to save a woman's life or in the case of rape. Although not explicitly permitted by law, abortions in the case of fetal malformation incompatible with neonatal life can be approved on a case by case basis by judicial discretion, a process that requires a lawyer's petition and statements by three physicians and a mental health professional. The law does not establish a legal gestational age limit; however, Ministry of Health guidelines recommend that abortions be conducted before 12 weeks gestation. Despite legal limitations, each year an estimated 1.4 million clandestine abortions are performed in Brazil resulting in some 300,000 hospitalizations for complications [1,2]. Unsafe abortion represents the third leading cause of maternal mortality, and deaths from unsafe abortion in Brazil comprise 12% of the maternal mortality ratio [1,3]. Brazil's abortion rate is high at 40.8 per 1,000 women, with approximately 31% of all pregnancies ending with an induced abortion [4,5].
In July 2004, a lone Brazilian federal judge issued a preliminary ruling that waived the requirement for court authorization for abortions in cases of fetuses with anencephaly (a fatal congenital defect that prevents formation of the brain), unleashing a fervent national debate that received substantial international media coverage. The Brazilian National Bishops' Conference lobbied hard for reversal of the ruling, whereas the National Confederation of Healthcare Workers (the governmental health sector union) pushed for its permanent acceptance. In October 2004, the full Brazilian Supreme Court convened and voted 7-4 to suspend the judge's solitary ruling until the full tribunal had the opportunity to deliberate and rule on the matter. As of November 2005, the issue still had not been settled and advocates continued to debate the moral, religious, and public health implications of liberalizing Brazilian abortion laws [6-8].
Although it has been the opinions of Brazilian judges and interest groups that have made headlines in this latest chapter of the country's ongoing abortion debate, physicians that play a crucial role in the accessibility and availability of safe and legal abortion. Their opinions on abortion, their willingness to perform an abortion, and their knowledge of abortion procedures directly affect whether their patients will be able to have access to safe and legal abortion. For this reason, it is important to understand their opinions and knowledge of abortion law, their knowledge of abortion procedures and their practices related to abortion.
Several studies have documented general public opinion of abortion in Brazil. A 1989 study by Meira and Ferraz examined medical and law student opinions of abortion law and found that almost half of the students surveyed thought that the law should allow for a legal abortion in more circumstances than currently permitted by Brazilian law. The majority of medical students thought that abortion should be legal in cases of rape (96%), risk to the woman's life or health (91%), congenital anomalies (86%), and mental incapacities (63%). A smaller percentage agreed with legalizing abortion in cases where the pregnant girl was less than 14 years old (36%) [9]. A survey of 1,456 women in a Southern county in Brazil revealed that 30% were in favor of legalizing abortion regardless of the circumstances, and those with higher education, higher family income and who had previously had an induced abortion were all more likely to support legalization of abortion[10]. In contrast to these women, a qualitative study of 71 Catholic, male college students in Brazil found that they generally had a negative opinion of abortion. When asked what they would do if a woman asked their advice regarding an induced abortion, 83% said they would counsel her not to abort [11].
A 2003 study in Brazil on attitudes towards abortion compared opinions of teenage women who had aborted, women who had considered abortion but ultimately did not abort, and women who did not abort. Initially, teens who had aborted and who had considered abortion were more tolerant of abortion than those who did not abort; however, their acceptance level decreased over time. Over the one-year follow-up period, the teens that did not consider abortion became more accepting of abortion. On average, across the three groups, 66% thought that an abortion was justified when the woman's life or health was in danger, 63% thought it was justified in the case of rape, and 47% felt that it was justified in the case of congenital anomalies [12].
In terms of abortion opinions of physicians, a small study conducted among 57 emergency room physicians in two São Paulo hospitals found that 31.5% demonstrated low knowledge of Brazilian abortion laws, and the vast majority were in favor of abortion in cases of rape (84%), risk to the mother's life (86%), and fetal malformation incompatible with life (82%) [13]. More recently, in 2003, Faúndes et al. surveyed a national sample of 4,261 Brazilian obstetricians-gynecologists (OB-GYNs) about abortion, asking participants whether they had helped their patients or relatives to have an abortion, and whether they themselves had had an abortion. The authors found that the respondents were progressively more accepting of legal abortion the closer they were to the person with the unwanted pregnancy: 41% of respondents had helped a patient to obtain an abortion, 49% had helped a relative to obtain an abortion, 78% of female physicians had obtained an abortion when they themselves had an unwanted pregnancy, and 80% of male physicians had helped their partners to obtain abortions. The vast majority of participants believed abortion should be legal if the woman's life is at risk (79%), if the pregnancy resulted from rape (80%), and in cases of fetal malformation (77%). Younger respondents were less supportive of abortion when confronted with unwanted pregnancies of patients or relatives, and twice as many physicians with no religious beliefs had helped patients or relatives to have an abortion compared to physicians to whom religion was very important. Nevertheless, when the physician herself or the male physician's partner had had an unwanted pregnancy, almost 70% of those to whom religion was very important had had an abortion [14,15].
The relevance of Brazilian physicians' abortion knowledge, attitudes, and practices extends far beyond the current legal limbo regarding abortion of anencephalic fetuses, and we seek to build on the Faundes et al. findings, some of which have not been published in an English-language journal [15]. We explored not only Brazilian OB-GYNs' knowledge and opinions on the legality of abortion in their country, but also their clinical experience and training (or lack thereof) in different abortion techniques [including medical abortion and manual vacuum aspiration (MVA)] and the different circumstances (e.g., in cases of risk to the woman's life or when the unwanted pregnancy is the result of a rape) under which they performed abortion. Faundes et al. documented OB-GYN abortion practices in terms of the relationship the provider held with the pregnant women, and we will expand on these findings by investigating the circumstances under which the pregnancies arose in those cases where OB-GYNs performed abortions. Finally, we will analyze the relationships between sociodemographic characteristics and experience in performing abortions on opinions on abortion law.
Methods
From December 2001 through September 2002, we mailed 1,500 questionnaires to a 10% random sample of OB-GYNs affiliated with the Brazilian Federation of Obstetricians and Gynecologists (FEBRASGO). At the time, FEBRASGO consisted of approximately 15,000 members. Our sample consisted of 10% of FEBRASGO members in each state, selected by assigning random digits to each member and then selecting a 10% random sample. In collaboration with FEBRASGO, we attempted to increase the response rate by using FEBRASGO stationery, publishing an advertisement in the FEBRASGO journal, sending reminder faxes, and offering a raffle. The Council's Institutional Review Board as well as FEBRASGO approved this study.
The anonymous questionnaire, which had undergone two rounds of pre-testing prior to fielding the study, contained 18 questions to assess respondents' knowledge of abortion law (specifically in cases of rape and life-threatening congenital malformations), their opinions of current abortion law in Brazil, their familiarity with various abortion procedures, and their experiences providing abortions. We also collected sociodemographic information, including sex, age, religion, and region of residence.
The primary outcome measure was opinion of abortion law. On the questionnaire, respondents were asked to choose one or more of the following circumstances in which he or she thought abortion should be legal in Brazil: life-threatening congenital malformation, rape, risk to a woman's life, risk to a woman's physical health, socioeconomic reasons, as an elective procedure, never legal, or under other circumstances. We classified respondents as conservative or liberal on abortion law based on their opinions of abortion law under the different circumstances. Conservative physicians were those who felt that abortion should only be legal under the circumstances codified in the current law (i.e., in rape cases or when the woman's life is in danger), or those physicians who thought the law should be more restrictive. Liberal physicians were those who felt that the current abortion law should be liberalized to allow for legal abortion under at least one other circumstance in addition to the two already permitted.
Using chi-square tests and crude logistic regression models, we determined which sociodemographic, knowledge-related or practice-related variables were associated with physician opinion. Following the bivariate analysis, we conducted a multivariate analysis of physicians' opinions of abortion law. In the initial multivariate analysis, we included variables that were associated with physician opinion of abortion law in the bivariate analysis at a p-value ≤ 0.10. We also included those variables that did not have a statistically significant relationship with opinion in the bivariate analysis, but were hypothesized to be associated with opinion based on the literature. For the final multivariate model, we used the Wald test to determine which variables were most highly associated with physician opinion of abortion law. In our final model, we include all variables that are significant at the p < 0.05 level as well as those that were marginally significant.
All analyses were conducted using SPSS statistical software version 10.0.6 and Stata, version 8.2.
Results
Out of 1,500 questionnaires mailed, we received 572 completed questionnaires (38% response rate). As shown in Table 1, the majority of the respondents were women (56%), between 26 and 45 years old (56%), and Catholic (72%). About 52% of the respondents lived in the Southeast Region of Brazil, which includes the states of Rio de Janeiro, Minas Gerais, São Paulo, and Espirito Santo. The geographical distribution of respondents was roughly equivalent to that of the FEBRASGO membership overall [14].
Table 1 Sociodemographic characteristics of OB-GYNs (n = 572)
Characteristic OB-GYNs n = 572a
n %
Gender
Male 252 44.1
Female 320 55.9
Age
26–35 119 21.5
36–45 199 35.9
46–55 159 28.7
56+ 77 13.9
Religion
Catholic 414 72.4
Evangelical 43 7.6
Other religion 61 10.8
Not religious 49 8.6
Region
South 97 17.0
Southeast 300 52.4
Northeast 87 15.2
North 17 3.0
Central 54 9.4
aMissing data: Age (18), religion (5), region (17).
Knowledge
Two hundred seventy-six physicians (48%) correctly identified that abortion is legal in Brazil to save the life of a woman and in the case of rape. Also, the large majority (70%) reported awareness that the Brazilian judicial system can grant court authorization of an abortion in cases of severe fetal malformations (Table 2). Considerable confusion existed over the Ministry of Health guidelines on the gestational age limit for abortion. One quarter of physicians did not know of such a gestational age limit guideline, while 40% thought it was 12 weeks and 30% thought it was 20 weeks. In terms of abortion procedures, 86% and 90% of physicians reported knowledge of MVA and dilation and curettage (D&C), respectively. A large majority (86%) also reported knowledge of misoprostol or other prostaglandins for abortion, while few physicians knew of the mifepristone and misoprostol regimen (37%) or the methotrexate and misoprostol regimen (27%).
Table 2 OB-GYN knowledge related to abortion (n = 572)
Knowledge OB-GYNs n = 572a
n %
Correct knowledge about abortion law
Yes 276 48.3
No 296 51.7
Familiar with MOH standards regarding rape
Yes 404 70.6
No 160 28.0
Aware that Brazilian judiciary system may authorize abortion in cases of serious fetal anomalies
Yes 404 70.6
No 144 25.2
Knowledge of gestational age limit for abortion
12 weeks 227 39.7
20 weeks 174 30.4
More than 20 weeks 5 < 1.0
Never legal 12 2.1
Do not know 142 24.8
Knowledge of procedures
Manual vacuum aspiration 493 86.2
Dilation and curettage 514 89.9
Hypertonic solutions (saline or urea) 140 24.5
Misoprostol* or other prostaglandins 493 86.2
Mifepristone + misoprostol 210 36.7
Methotrexate + misoprotol 153 26.7
Other 4 0.70
aMissing data: Familiar with MOH standards regarding rape (8), aware that Brazilian judiciary system may authorize abortion in cases of serious fetal anomalies (24), knowledge of gestational age limit for abortion according to law (12)
* Otherwise known as Cytotec
Opinion
Table 3 contrasts the percentage of physicians who believed that abortion is currently legal in Brazil, according to different circumstances, with the percentage of physicians who believed abortion should be legal. In every circumstance except for rape, more physicians believed that abortion should be legal than believed abortion is currently legal. In the case of rape, almost all (93%) correctly identified that abortion is legal and a lesser percentage (85%) thought that it should be legal. A sizeable difference emerged in the case of severe fetal malformations, where 36% of physicians incorrectly believed that it was a legal case for abortion, while 89% thought that abortion should be legal in this case. We observed a similarly large difference when the woman's health is at risk; 7% incorrectly thought that the law currently allowed for an abortion in this case, whereas 30% of physicians thought that abortion should be legal in this case.
Table 3 OB-GYN knowledge and opinion of Brazilian abortion law by circumstance (n = 572)
Circumstance Believe abortion is currently legal Believe abortion should be legal
n % n %
Pregnancy as a result of rape* 529 92.5 488 85.3
Risk to the woman's life* 453 79.2 493 86.2
Severe fetal malformation 205 35.8 506 88.5
Risk to the woman's health 38 6.6 180 31.5
Socioeconomic reasons 3 < 1.0 74 12.9
When the woman chooses 1 < 1.0 76 13.3
Never 12 2.1 23 4.0
* At the time of data collection, abortion was legal under this circumstance
Overall, we found that only 3% of physicians agreed with the current law that allows for abortion only in cases of rape or to save the woman's life. Most physicians (77%) thought the law should be liberalized to allow for a legal abortion under more circumstances and 17% thought the law should be more restrictive. The overwhelming majority (95%) expressed their support for public funding of abortion services (Table 4).
Table 4 OB-GYN opinion about abortion (n = 572)
Opinion OB-GYNs n = 572a
n %
Believe abortion should only be legal in cases that are currently allowed by law*
Yes 16 2.8
No 556 97.2
Believe abortion law should be more restrictive than the current law*
Yes 99 17.3
No 473 82.7
Believe abortion law should be more liberal than the current law*
Yes 443 77.4
No 126 22.0
Support public funding for abortion services
Yes 545 95.3
No 24 4.2
aMissing data: Believe abortion law should be more liberal than the current law (3), support public funding for abortion services (3)
* At the time of data collection, abortion was legal in cases of rape or to save the life of the mother
Practices
In addition to sharing their knowledge and opinions regarding abortion law, physicians also reported on their abortion-related practices (Table 5). Nearly 70% had never received any training on abortion procedures. Among the 33% of respondents that had ever performed an abortion, the most common procedures were D&C (60%) and the use of misoprostol or other prostaglandins (68%) for medical abortion. While most of the respondents (73%) had performed an abortion within the first 12 gestational weeks, there was also a significant number (44%) that had performed an abortion between 13 to 20 gestational weeks. Fifty-three percent had performed an abortion in the case of severe fetal malformations, 25% had performed an abortion to save a woman's life, and 19% had performed one in a case of rape.
Table 5 Practices related to abortion among OB-GYNs (n = 572)
Practice OB-GYNs n = 572a
n %
Ever received training
Yes 176 30.8
No 393 68.7
Ever performed an abortion
Yes 188 32.9
No 380 66.4
Number of abortions performed in the last year*
1–5 90 47.9
6–20 10 5.3
More than 20 5 2.7
None in the last year 83 44.1
Circumstances under which abortions were performed*
Rape 36 19.1
Risk to the mother's life 47 25.0
Severe fetal malformation 100 53.2
Risk to the mother's health 9 4.8
Socioeconomic reasons 3 1.6
Elective 18 9.6
Other 38 20.2
Procedures used*
Manual vacuum aspiration 18 9.6
Dilation and curettage 112 59.6
Hypertonic solutions (saline or urea) 6 3.2
Misoprostol* or other prostaglandins 129 68.6
Mifepristone + misoprostol 4 2.1
Methotrexate + misoprotol 5 2.6
Other 11 5.9
Gestational age when performed abortion*
Before 12 weeks 137 72.9
Between 13–20 weeks 83 44.1
a Missing data: Ever received training (3), ever performed an abortion (4)
* Only physicians who reported having performed an abortion answered this question (n = 188); participants could choose more than one option
We encountered several statistically significant relationships between physicians' sociodemographic characteristics and abortion-related practices with their opinion of abortion law (Table 6). In a bivariate analysis, physicians who felt the law should be more liberal were more likely to have correct knowledge of abortion law, to be familiar with the abortion law regarding severe fetal malformations, and to support public funding for abortion services (p < 0.10). Catholic physicians and Evangelical physicians were not significantly different in their abortion opinions from those who reported that they were not religious; however, physicians who reported having "other" religious affiliation had lower odds of thinking the abortion laws should be more liberal. These variables were all included in an initial multivariate analysis of physician opinion, in addition to region of residence and whether or not they had ever performed an abortion. The latter variables were included due to their hypothesized association with physician opinion and their importance in the literature. In the final model, we combined Catholic and Evangelical physicians into a single group and compared them with non-religious respondents and respondents reporting "other" as their religion. We found that Catholic and Evangelical physicians were not significantly different in their abortion opinions from physicians who reported no religious affiliation. In contrast, physicians whose religious affiliation was "other" had a decreased odds of thinking abortion law should be more liberal compared to physicians with no religious affiliation (OR 0.27 (95% CI: 0.10, 0.74). In addition, physicians who had correct knowledge regarding the status of abortion law in Brazil had a 46% higher odds of favoring a more liberal abortion law compared to those who did not have correct knowledge (OR 1.46 (95% CI: 0.96, 2.22), but this finding was not statistically significant. Finally, physicians who were in favor of public funding for legal abortion services had six times the odds of favoring a more liberal abortion law compared to those who were opposed to public funding of legal abortions (OR 6.01 (95% CI: 2.53, 14.28); however, it must be taken into account that only a small number (24) had said that they opposed public funding for legal abortions. Region of residence, correct knowledge of the law regarding abortions in cases of severe fetal malformations, and experience performing an abortion were not significantly related to physician opinion in the multivariate analysis and thus not included in the final model.
Table 6 OB-GYN opinions of abortion law in Brazil according to selected demographic characteristics, knowledge, and practices (n = 572)
Characteristic Law should be more liberal Odds Ratios and 95% Confidence Intervals
N 443 % 77.9 OR 95% CI
Gender
Female 245 77.0 0.90 (0.60, 1.34)
Male 198 78.9 1 (ref)
Age
26–35 97 81.5 1 (ref)
36–45 159 79.9 0.90 (0.51, 1.61)
46–55 119 75.8 0.71 (0.39, 1.28)
56+ 60 79.0 0.85 (0.41, 1.74)
Religion
Catholic 328 79.6 0.54 (0.22, 1.32)
Evangelical 31 72.1 0.36 (0.12, 1.06)
Other 38 62.3 0.23 (0.08, 0.63)
Not religious 43 87.8 1 (ref)
Region
South 76 78.4 1 (ref)
Southeast 234 78.5 1.01 (0.58, 1.76)
Central 45 83.3 1.38 (0.58, 3.27)
North 12 70.6 0.66 (0.21, 2.09)
Northeast 61 70.9 0.67 (0.34, 1.32)
Knowledge of abortion law
Yes 217 74.1 1.58 (1.06, 2.37)
No 226 81.9 1 (ref)
Familiar with MOH standards on sexual violence and abortion
Yes 317 78.7 1.15 (0.74, 1.77)
No 122 76.3 1 (ref)
Knowledge of gestational age limits on abortion in cases of rape
Yes 264 79.8 0.76 (0.51, 1.13)
No 170 74.9 1 (ref)
Surgical methods known
Yes 431 78.6 2.05 (0.67, 6.22)
No 9 64.3 1 (ref)
Medical methods known
Yes 402 78.8 1.37 (0.72, 2.62)
No 38 73.1 1 (ref)
Knowledge of law regarding fetal malformation
Yes 325 80.6 1.67 (1.08, 2.60)
No 102 71.3 1 (ref)
Received training
Yes 137 77.8 0.99 (0.65, 1.52)
No 305 78.0 1 (ref)
Support public funding for abortion services
Yes 434 79.9 6.64 (2.83, 15.57)
No 9 37.5 1 (ref)
Ever performed an abortion
Yes 150 80.2 1.21 (0.78, 1.86)
No 292 77.0 1 (ref)
Discussion
Knowledge
Although abortion is legal in Brazil to save a woman's life and in cases of rape, Brazilian women have limited access to safe abortion services. Two recent articles ascribed a portion of this limited access to legal abortion in Brazil to a lack of knowledge of abortion law in the general population [16,17]. Indeed, in our survey of Brazilian OB-GYNs, about half of the physician respondents could not correctly identify the Brazilian abortion law. Additionally, a significant number of physicians thought that abortion was legal in the case of severe fetal malformations when in fact, at the time of data collection, women were required to solicit permission from the judicial system for an abortion in these cases. The large number of physicians who were not familiar with Brazilian abortion law is alarming because these OB-GYNs will be unable to give accurate information to their patients. The participants' confusion may now be exacerbated by the recent turmoil surrounding abortion of anencephalic fetuses. Our findings are similar to those of Faundes et al., who observed that two-thirds of Brazilian OB-GYNs wrongly believed that a judicial order is required to obtain a legal abortion and only 27% knew that the woman needed to make a written request to obtain a legal abortion [15].
The majority of OB-GYNs surveyed reported familiarity with manual vacuum aspiration (MVA) and the dilation and curettage (D&C) procedures. Additionally, a large majority of physicians knew of the possibility of using misoprostol or other prostaglandins to induce abortion. Misoprostol, which is marketed under the name Cytotec, has received significant press coverage in Brazil as the drug's worldwide notoriety as an abortifacient first began in Brazil in 1988 [18]. However, fewer OB-GYNs surveyed knew of the use of misoprostol in combination with either mifepristone or methotrexate, which are more successful regimens for induced abortion [19-21].
Opinion
In terms of opinions on abortion law, very few physicians agreed with the current law, mostly due to the fact that an overwhelming majority thought that abortion should be legal in the case of severe fetal malformations (i.e., they felt abortion should be legal without needing judicial authorization). As an illustration of this support, out of the 188 physicians that reported having performed an abortion, over half of them said they had performed an abortion because of severe fetal malformations. In addition, there was some support for abortion being legal when the woman's health is at risk, for socioeconomic reasons, and when a woman chooses.
Overall, OB-GYNs who had correct knowledge of abortion law, who reported to be Catholic, Evangelical, or not religious (as compared to "other" religion), and who supported public funding of legal abortion services were more likely to be in favor of a more liberal abortion law. A limitation of our study in terms of this particular finding is that respondents did not specify what their religion was if it fell under the "other" category, so we cannot hypothesize about why these participants would have more conservative attitudes about abortion. Finally, contrary to a priori hypotheses, physicians who had experience in conducting abortions were not more likely to support abortion being legal under additional circumstances than those who had not performed an abortion.
Practices
About a third of OB-GYN respondents reported ever having performed an abortion. Thirty-six (19%) of those respondents who had ever performed an abortion reported doing so in the case of a rape, and 47 (25%) had performed an abortion in the case of risk to the woman's life, both of which were circumstances under which abortions were legally permitted. Nevertheless, while not specifically allowed by the law at the time of data collection, 100 (53%) of the physicians who reported ever having performed an abortion said that the abortion was a case of severe fetal malformation. It is striking that, among those respondents who had ever performed an abortion, most had done so under circumstances not explicitly permitted by law at the time of data collection.
D&C and misoprostol alone (or other prostaglandins) were the two most commonly used procedures for inducing abortion among the OB-GYNs surveyed. Very few physicians reported using MVA, despite wide recognition of its effectiveness and its increased safety and cost effectiveness over D&C [22,23]. Similarly, few physicians reported using misoprostol in combination with either mifepristone or methotrexate, even though these are more effective regimens than using misoprostol alone [19-21].
Conclusion
The majority of OB-GYNs in our study and in the previous Faundes et al. study agreed that abortion should be legal when the pregnancy endangers the life of the woman or is a result of a rape (in those circumstances abortion is permitted under current Brazilian law)[14,15]. However, among those OB-GYNs with previous experience performing abortions, more than half of them had done so in cases of fetal malformation, a circumstance that was not explicitly permitted under current abortion law. Not only do these findings underscore the widespread confusion and misperceptions surrounding abortion law, but they also provide powerful evidence of the fact that abortions are being performed by OB-GYNs even in cases that are technically illegal. This information is particularly valuable to those advocating a liberalization of abortion law to permit legal abortions in cases of fetal malformation. This study also demonstrates the need for training in surgical and medical abortion techniques for Brazilian OB-GYNs. MVA is not as well known as D&C, even though the former is both more effective and less expensive than the latter. Likewise, training on medical abortion is needed to educate physicians on the benefits of using misoprostol in conjunction with mifepristone or methotrexate. Educational efforts, ideally beginning with medical trainees in universities, should focus on introducing these safer, less costly abortion techniques in Brazil. In conclusion, we hope this study can inform educational campaigns for Brazilian OB-GYNs in order to clarify misperceptions of abortion law and increase their technical capacity to provide safe, legal abortions.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
JD and SGG conceived of the study and participated in its design and coordination. LAG, SGG, and EAY performed data analysis and helped draft the manuscript. All authors read and approved the final manuscript.
Table 7 Adjusted odds ratios of the likelihood of thinking that the abortion law should be more liberal (n = 572)
Characteristic Odds Ratio 95%CI p-value
Religion
Catholic/Evangelical 0.63 (0.27, 1.52) 0.301
Other religion 0.27 (0.10, 0.74) 0.011
Not religious 1 (ref)
Correct knowledge of abortion law
Yes 1.46 (0.96, 2.22) 0.077
No 1 (ref)
Support public funding
Yes 6.01 (2.53, 14.28) < .001
No 1 (ref)
Acknowledgements
This study was supported in part by the John D. and Catherine T. MacArthur Foundation and the William and Flora Hewlett Foundation. We thank Davida Becker, Loren Galvao, and Christine Ricardo for their help on this study.
==== Refs
Guedes AC Abortion in Brazil: legislation, reality and options Reprod Health Matters 2000 8 66 76 11424252 10.1016/S0968-8080(00)90188-5
Singh S Wulf D Estimated levels of induced abortion in six Latin American countries International Family Planning Perspectives 1994 20 4 13
Pan American Health Organization Evaluación del plan de acción regional para la reducción de la mortalidad materna 1996
Henshaw SK Singh S Haas T The incidence of abortion worldwide Int Fam Plann Persp 1999 25 S30 8 14627053
Ipas Brazil Ipas Brazil report
Hall KG Brazilian court slams door on easing of abortion rules The Miami Herald 2004 1A
Osava M Abortion of anencephalic fetuses fires debate Inter Press News Agency 2004
Lobo I Brazil: Catholic Church goes to court against abortion Brazzil Magazine 2004
Meira AR Ferraz FR [Abortion liberation: the opinion of medical and law students, Sao Paulo, Brazil] Rev Saude Publica 1989 23 465 472 2641838
Cesar JA Gomes G Horta BL de Oliveira AK Saraiva AK Pardo DO Silva LM Rodghiero CL Gross MR [Women's opinion on abortion legalization in a middle size county in southern Brazil] Rev Saude Publica 1997 31 566 571 9629711
de Brito RS Almeida MS Enders BC [The knowledge of male university students on induced abortion] R Bras Enferm 2000 53 173 182
Bailey PE Bruno ZV Bezerra MF Queiros I Oliveira CM Adolescents' decision-making and attitudes towards abortion in north-east Brazil J Biosoc Sci 2003 35 71 82 12537157 10.1017/S0021932003000713
Loureiro DC Vieira EM [Knowledge and opinions concerning legal and ethical issues with abortion among physicians working in emergency wards in Ribeirao Preto, Sao Paulo, Brazil] Cad Saude Publica 2004 20 679 688 15263978
Faundes A Duarte GA Neto JA de Sousa MH The closer you are, the better you understand: the reaction of Brazilian obstetrician-gynaecologists to unwanted pregnancy Reprod Health Matters 2004 12 47 56 15938157 10.1016/S0968-8080(04)24011-3
Faundes A Alves G Andalaft J Olivatto AE Martins R [Knowledge, opinion and attitudes of Brazilian gynecologists and obstetricians regarding induction of abortion] Rev Bras Ginecol Obstet 2004 26 89 96
Faundes A Leocadio E Andalaft J Making legal abortion accessible in Brazil Reprod Health Matters 2002 10 120 127 12369314 10.1016/S0968-8080(02)00017-4
Villela WV Araujo MJ Making legal abortion available in Brazil: partnerships in practice Reprod Health Matters 2000 8 77 82 11424253 10.1016/S0968-8080(00)90189-7
Barbosa RM Arilha M The Brazilian experience with Cytotec Stud Fam Plann 1993 24 236 240 8212093
Jain JK Dutton C Harwood B Meckstroth KR Mishell DR A prospective randomized, double-blinded, placebo-controlled trial comparing mifepristone and vaginal misoprostol to vaginal misoprostol alone for elective termination of early pregnancy Hum Reprod 2002 17 1477 1482 12042265 10.1093/humrep/17.6.1477
Kahn JG Becker BJ MacIsaa L Amory JK Neuhaus J Olkin I Creinin MD The efficacy of medical abortion: a meta-analysis Contraception 2000 61 29 40 10745067 10.1016/S0010-7824(99)00115-8
Blanchard K Winikoff B Ellertson C Misoprostol used alone for the termination of early pregnancy. A review of the evidence Contraception 1999 59 209 217 10457864 10.1016/S0010-7824(99)00029-3
Ladipo OA Preventing and managing complications of induced abortion in Third World countries Suppl Int J Gynecol Obstet 1989 3 21 28 2686704 10.1016/0020-7292(89)90099-4
Kizza AP Rogo KO Assessment of the manual vacuum aspiration (MVA) equipment in the management of incomplete abortion East Afr Med J 1990 67 812 822 2076683
|
16288647
|
PMC1308861
|
CC BY
|
2021-01-04 16:38:15
|
no
|
Reprod Health. 2005 Nov 15; 2:10
|
utf-8
|
Reprod Health
| 2,005 |
10.1186/1742-4755-2-10
|
oa_comm
|
==== Front
RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-671627448210.1186/1742-4690-2-67ResearchEvolution of subtype C HIV-1 Env in a slowly progressing Zambian infant Zhang Hong [email protected] Federico [email protected] Jun [email protected] Xiang [email protected] Chipepo [email protected] Ruth [email protected] John T [email protected] Guillermo [email protected] Charles [email protected] Nebraska Center for Virology, University of Nebraska, Lincoln, NE, USA2 The School of Biological Sciences, University of Nebraska, Lincoln, NE, USA3 Department of Pediatrics, University Teaching Hospital, Lusaka, Zambia4 Dana-Farber Cancer Institute, Harvard Medical School, Boston, MA, USA2005 7 11 2005 2 67 67 30 6 2005 7 11 2005 Copyright © 2005 Zhang et al; licensee BioMed Central Ltd.2005Zhang et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Given the high prevalence of mother to child infection, the development of a better understanding of African subtype C HIV-1 transmission and natural evolution is of significant importance. In this study, we genotypically and phenotypically characterized subtype C viruses isolated over a 67-month follow-up period from an in utero-infected Zambian infant. Changes in genotype and phenotype were correlated to alterations of the host humoral immune response.
Results
A comparison of baseline maternal and infant samples indicated that the infant sequences are monophyletic and contain a fraction of the diversity observed in the mother. This finding suggests that selective transmission occurred from mother to child. Peaks in infant HIV-1 Env genetic diversity and divergence were noted at 48 months, but were not correlated with changes in co-receptor usage or syncytia phenotype. Phylogenetic analyses revealed an accumulation of mutations over time, as well as the reappearance of ancestral lineages. In the infant C2-V4 region of Env, neither the median number of putative N-glycosylation sites or median sequence length showed consistent increases over time. The infant possessed neutralizing antibodies at birth, but these decreased in effectiveness or quantity with time. De novo humoral responses were detected in the child after 12 months, and corresponded with an increase in Env diversity.
Conclusion
Our study demonstrates a correlation between HIV-1 Env evolution and the humoral immune response. There was an increase in genetic diversification in the infant viral sequences after 12 months, which coincided with increases in neutralizing antibody titers. In addition, episodes of viral growth and successive immune reactions in the first 5–6 years were observed in this slow progressor infant with delayed onset of AIDS. Whether this pattern is typical of slow progressing subtype C HIV-1 infected infant needs to be further substantiated.
==== Body
Background
Subtype C human immunodeficiency virus type 1 (HIV-1) accounts for over 56% of HIV-1 infections [1-3]. Globally, HIV-1 infection is one of the leading causes of childhood morbidity and mortality. HIV-1 infected children account for 20% of all HIV-1 related deaths; 7% of individuals living with HIV-1 infection, and 16% of new HIV-1 infections annually [4]. In sub-Saharan Africa, HIV-1 subtype C is responsible for approximately 50% of infections and a significant number of infections are in infants and children. Transmission of HIV-1 from infected mothers to their infants is the primary mode of HIV-1 infection in children and can occur in utero, intrapartum, or postnatally through breast milk. The use of antiretroviral regimens has successfully reduced the rate of HIV-1 infection in infants in the developed world to approximately 1%; nevertheless, such regimens have only recently become available in many of the developing nations where HIV-1 mother to child transmission (MTCT) is most significant [5].
HIV-1 MTCT is complex, and its determinants are not completely understood. Several factors, including high maternal viral load, maternal env gene homogeneity, and rapid viral replication kinetics, have been correlated with perinatal HIV-1 transmission [6-8]. In addition, advanced maternal disease status, lack of drug therapy, and lack of breast-feeding alternatives contribute to increased MTCT [9]. Moreover, several studies have demonstrated the transmission of minor [9-12], major [9,11], and multiple [9,13,14] HIV-1 genotypes from mother to infant. Our understanding of perinatal transmission and disease progression in infants is mainly derived from studies of subtype B infected individuals. The applicability of such findings to other subtypes remains to be substantiated.
The natural history of subtype C HIV-1 infection has not been extensively studied in children. It is known that infant disease survival times are considerably shorter than those of HIV-infected adults, and that without treatment, most HIV-1 infected African children die before their third birthday [15]. Given the expanding distribution of subtype C infections, a complete understanding of virus transmission and natural evolution is increasingly important.
HIV-1 transmission is, in part, a function of the receptor binding by the envelope glycoprotein (Env) that mediates virus-cell fusion. Alteration of Env has been linked to expanded host range, alternative co-receptor usage and in vitro syncytium induction and associated with viral pathogenesis and disease progression [16-25]. Accumulating evidence suggests that subtype C Env displays biological properties, such as near-exclusive CCR-5 utilization, that distinguish it from other subtypes. In addition, the subtype C Env glycoprotein, third variable region (V3) is more conserved than the previously defined "constant" regions [26,27]. Whether differences in cellular tropism, transmission and pathogenetic outcome observed between subtype C and other subtypes correlate with the Env glycoprotein biological or genetic properties need to be examined. In addition, whether there exist differences in Env evolution in infected children based on viral subtype, remains to be determined. Recently it has been suggested that particular changes in env in Zambian adults correlated with heterosexual transmission. Viruses with shorter Env length, and fewer putative N-linked glycosylation sites (PNGS) were suggested to be more susceptible to neutralizing antibodies, yet more efficient at transmission [28]. Similar correlates have not been reported for transmission to children.
In the present study, we investigated the longitudinal variation of the viruses in a subtype C HIV-1 infected Zambian mother/infant pair (MIP 1157). This pair was antiretroviral therapy naïve over a six-year follow-up period. The extended follow-up enabled us to examine the interplay between humoral immune selection and virus evolution. We describe changes in the infant Env C2-V4 region over the follow-up period, and correlate these changes with alterations in viral phenotype and host humoral immune response. Our findings indicate that genetic diversification in the infant Env gene increased after 12 months, and is correlated with increases in neutralizing antibody titers.
Results
HIV-1 infected mother-infant pair
We characterized HIV-1 transmission and longitudinal evolution of the HIV-1 envelope glycoprotein in a Zambian mother and infant pair (MIP 1157) for more than 6 years. The mother and child are anti-retroviral naïve and remain clinically asymptomatic. Infant 1157 was infected in utero since HIV-1 sequences were detected by DNA PCR of infant blood samples collected at birth. The baby was delivered naturally, healthy and with normal birth weight, and was breast-fed until 20 months of age. The child remains clinically asymptomatic throughout the follow-up study period and his CD4 counts was 658 cells/μl at 6 years old. The child has been evaluated at the study clinic where blood specimens were collected every 6 months for the first 24 months and at 12-month intervals thereafter. The prolonged survival of this infected child is unusual since most untreated HIV-1 infected African children do not survive beyond the first three years of life. The extended follow-up of infant1157 provided us with an opportunity to investigate correlates of virus transmission in the Env glycoprotein and to track genetic variation and evolution of this gene over time.
MIP1157 viruses use CCR5 as co-receptor and belong to subtype C
All viral isolates recovered from MIP 1157 replicated efficiently in PBMC and monocyte-derived macrophages (MDM), but failed to grow in MT-2 and C8166 T-cell lines. Viral isolates did not induce syncytia in infected PBMC and MDM. We evaluated viral co-receptor usage in cell lines that co-express CD4 with a single co-receptor. All isolates failed to infect CXCR4-expressing CEMx174-GFP cells, and similarly, none of the viruses grew in cells expressing only CCR3 (HOS-CD4-CCR3) (data not shown). In addition, 1157 viruses failed to replicate in PBMC homozygous for the Δ32 deletion variant of CCR5. In contrast, cells expressing normal CCR5 and CD4 (GHOST-CD4-CCR5) were readily infected, suggesting that 1157 HIV-1 isolates primarily use CCR5 as co-receptor (data not shown). This is in agreement with infectivity assays demonstrating that only primary PBMC and MDM support viral growth. Phylogenetic analyses clustered all 1157 env sequences with subtype C.
Transmission pattern
Viral env sequences from both the mother and infant at birth were analyzed to examine the genealogical pattern of perinatal transmission. Infant birth samples were monophyletic relative to the mother in all phylogenetic analyses (Bayesian [BA], maximum likelihood [ML] and neighbor joining [NJ]). In all cases, phylogenetic trees support the concept of a restricted pattern of transmission, where a subset of the maternal quasispecies was passed into the child (Figure 1). As would be expected in a restricted transmission, genetic variation in the HIV-1 Env gene is lower in infant birth sequences than in maternal sequences from the same timepoint (Table 1). The mean number of nucleotide substitutions within the mother's env sequences at birth was 3.2, compared to 1.67 in the infant (Table 1), and the mean number of amino acid differences was 2 in the mother and 1 in the infant. These findings from phylogenetic and diversity analyses indicate that the infant possesses a subset of the maternal diversity at the time of birth.
Figure 1 Phylogenetic relationships between mother (thin) and infant (thick) samples collected at birth. Majority rule consensus from a Bayesian analysis (BA) run for 5 × 106 generations, sampled every 1000. The last 3000 trees were used to build the consensus. Posterior probabilities are next to the relevant nodes.
Table 1 Viral variations in the different mother and infant populations
Sample n H Nuc AA PNGS L
Infant at birth 48 26 2 (0 – 7) 1 (0 – 5) 15 (14 – 15) 183 (183 – 183)
Infant 6 months 29 23 4 (0 – 11) 3 (0 – 9) 14 (13 – 15) 183 (175 – 183)
Infant 12 months 27 24 6 (0 – 15) 4 (0 – 11) 13 (13 – 15) 174 (174 – 183)
Infant 18 months 51 38 15 (0 – 29) 11 (0 – 23) 14 (13 – 15) 179 (175 – 183)
Infant 24 months 37 36 14 (1 – 21) 11 (0 – 18) 12 (11 – 16) 177 (174 – 183)
Infant 29 months 28 27 16 (1 – 26) 12 (0 – 19) 12.5 (11 – 15) 182 (173 – 183)
Infant 36 months 26 24 13 (0 – 27) 8 (0 – 19) 12 (10 – 14) 176 (173 – 183)
Infant 48 months 25 25 25 (2 – 37) 16 (2 – 26) 13 (12 – 14) 182 (176 – 185)
Infant 67 months 32 24 35 (0 – 57) 24 (0 – 37) 13 (12 – 15) 183 (180 – 185)
Mother at delivery 26 20 5 (1 – 10) 3 (0 – 7) 15 (14 – 15) 183 (183 – 183)
Mother 12 months 32 31 8 (1 – 23) 5 (0 – 13) 15 (13 – 15) 183 (183 – 183)
Mother 18 months 33 17 5 (0 – 13) 2 (0 – 9) 15 (13 – 15) 183 (183 – 183)
Mother 24 months 32 18 2 (0 – 7) 1 (0 – 5) 14 (13 – 14) 183 (183 – 183)
Number of samples per timepoint (n); number of unique haplotypes (H); number of nucleotide differences (nuc) as median (min-max); number of amino acid differences (AA) as median (min-max); number of putative N linked glycosylation sites (PGNS) as median (min-max), and sequence length in codons (L) as median (min – max).
Longitudinal variation in env sequences
Given the lack of diversity in Env from the infant birth sample, and the extended survival of the child in the absence of antiretroviral therapy, it was of significant interest to investigate evolution of the Env gene over time. Since antiretrovirals were unavailable, the primary selective pressures acting on Env from infant 1157 were maintenance of replication and immune surveillance. Population-level changes in the genetic make-up of the quasispecies within the infant were followed by measuring genetic divergence and genetic diversity over time. Genetic divergence measures the number of differences from each contemporaneous set of sequences relative to the baseline population, whereas genetic diversity is an estimate of effective population size based on the average number of pair-wise differences within each set of contemporaneous sequences. The genetic diversity and genetic divergence of the infant Env C2-V4 region increased up to 48 months, but subsequently decreased or leveled off (Figure 2).
Figure 2 Changes in genetic divergence and diversity over time for the infant 1157. Panel A, Genetic divergence, as the average number of changes between each time point and the initial population, collected at birth. Panel B, Genetic diversity, as θπ, calculated from the average number of nucleotide differences within a given time point, which correlates with effective population size. The first plot describes the amount of change relative to the initial population and the second one describes the amount of variation within a time point.
Changes in Env genetic divergence and diversity, and in particular, the replacement of lineages over time (correlated with the stabilization of diversity and divergence), become evident when visualized in a phylogenetic tree. We constructed phylogenetic trees using NJ, ML and BA. All methods yielded similar results and only the NJ result is shown. There is an association between time of collection and sequence change (longer branches denote more changes) as early time point sequences appear on short branches, scattered at the base of the tree, while later sequences appear on long branches (Figure 3). Samples collected at 67 months are grouped into 6 different lineages, three that are closely associated with 48-month sequences, and three that are associated with sequences from earlier lineages. These would indicate that viral lineages persist in the infant and reappear at later times, e.g. some sequences collected at 67 months are closely related to sequences collected at 12, 18, and 48 months (see arrows in Figure 3). Alternatively, the virus may be selected to recreate those previous lineages as the immune pressure on particular epitopes in Env wanes.
Figure 3 Neighbor-joining (NJ) tree describing phylogenetic relationships between mother (black) and infant (colors) samples collected from all timepoints, using a GTR model of nucleotide substitution. Labels indicate the time of collection (i. e.: i06 corresponds to sequences from the infant collected 6 months after birth).
The temporally dependent lengthening of branches seen in phylogentic trees from BA, ML and NJ analyses was similar to the idealized shape expected under continual selection. As an estimate of the relative strength of selective pressure we calculated the ratio of non-synonymous (dN) to synonymous (dS) changes (dN/dS) for each timepoint. We observed a high ratio of non-synomymous to synonymous substitutions over time as estimated by ML in PAML (Figure 4). Estimates of the overall dN/dS ratio in the infant ranged from 0.42 (i00 m) to 1.36 (i24 m). We next calculated the number of synonymous and non-synonymous substitutions per codon for each contemporaneous set of sequences to assess how selective pressure was distributed along the region of Env sequenced. Non-synonymous variation was evenly distributed in the mother and infant throughout the fragment at baseline (Figure 4). As time progressed, the number of non-synonymous changes increased in the infant Env (Figure 4), but not in the mother (data not shown). A comparison across timepoints indicates that non-synonymous variation concentrated on the first portion of the constant region 2 (C2), the first portion of the constant region 3 (C3), and the terminal portion of the variable loop 4 (V4) (Figure 4). The overall high values of dN/dS, indicated by the relative amounts of red and green in the different panels of figure 4, and tree shape (Figure 3) suggest that positive Darwinian selection is playing a strong role in shaping molecular evolution in these samples.
Figure 4 Synonymous and non-synonymous amino acids variation along the HIV- 1 Env constant region 2 (C2), variable loop 3 (V3), constant region 3 (C3), and variable loop 4 (V4). Results are presented for maternal and infant samples collected at birth, as well as for infant samples collected from 6 to 67 months. Synonymous (green) and non-synonymous (red) changes per position for each sequence set were estimated in Datamonkey. The number and position of putative N-linked glycosylation sites (PNGS) (N × T/S) was estimated in N-GlycoSite . Within each set of contemporaneous sequences, constant PNGS are indicated in purple, and variable ones with blue (with their frequency in the blue outlined box). The overall rate of non-synonymous to synonymous substitutions (dN/dS) was estimated in PAML. N: number of sequences for each timepoint.
A recent report suggested that subtype C viruses transmitted between members of Zambian discordant couples possess envelope glycoproteins that are under-glycosylated, neutralization sensitive and contain short loop structures [28]. To explore the potential role of specific sequence characteristics in virus transmission between mother and child, we compared the sequence length polymorphism and variation in the number of PNGS for baseline maternal and infant Env C2-V4 sequences. There are 15 PNGS in this region of 1157 Env. Maternal and infant baseline sequences are all of the same length, and showed little variation in the PNGS (Table 1 and Figure 4). In the mother, there were 4 sequences, out of 26, that lost a PNGS, and the position at which this site was ablated was not conserved among any of the four. In the infant, 6 of 48 sequences lost a single PNGS, but in parallel with the mother, there was no conservation in the position of that loss. Moreover, only one variable PNGS was shared between the mother and the infant.
A similar evaluation of the C2-V4 length polymorphism and PNGS alteration was carried out on subsequent infant samples to assess the longitudinal variation in these two parameters (Table 1 and Figure 4). Length polymorphism was only observed in infant sequences where putative insertions and/or deletions occur in a subset of sequences at amino-acid positions 106–109 and 166–180. Maternal sequences remained of constant length, 183 amino acids, throughout the follow-up. All transmitted sequences in the infant were initially of the same length (also 183 amino acids). Length polymorphism in the region spanning amino acids 166–180 appeared 6 months postpartum, whereas polymorphism in the region spanning 106–109 was first observed at 12 months. The longest sequences, isolated at 48 and 67 months, were 185 amino acids in length, whereas the shortest sequences, 173 amino acids, were isolated at 29 and 36 months. All infant PNGS present at baseline remain present in a fraction of sequences from subsequent timepoints; however, only 3 sites remained fixed over the entire course of infection. The largest PNGS variation was observed at positions 7, 104, and 177, which oscillate between high and low prevalence (Figure 4, months 24, 36 and 48; position 7). In addition, there are 2 sites gained, one at position 72 at 6 months, and the other at position 109 at 12 months. Both of these polymorphisms are low in frequency and both are adjacent to another PNGS.
Replication Kinetics
In order to determine whether there are differences in the rates of replication between early and late viral isolates, the replication kinetics of the infant isolates from 6, 12 and 48-month in primary PBMC were determined by measuring the accumulation of RT units in supernatant over time. All the viral isolates displayed similar replication kinetics with a steady increase during the first 3 days of incubation and peaked by day 3(Figure 5). The RT units dropped after 3 days and remained relatively stable for the duration of the experiment (Figure 5). In addition, the similar replication kinetics of these viral isolates was also observed in MDM (data not shown).
Figure 5 Replication kinetics of 1157 infant viral isolates obtained at 6, 12 and 48-month follow-ups were determined in PBMC culture supernatant by measuring RT units. The laboratory viral strain 128 A was used as control. Each 100 TCID50 viral inoculum was added to 6 × 106 PHA-stimulated PBMC from a HIV-1 seronegative blood donor. RT units were measured in culture supernatant at day 0, 1, 3, 5, 7, 9, 11, 13, 15 post-infection.
Neutralization of infant HIV-1 isolates
To determine whether Env evolution correlated with the development of infant anti- HIV-1 humoral immunity, we analyzed neutralization of infant 6-month, 12-month and 48-month viral isolates by contemporaneous and non-contemporaneous plasma (Figure 6). The neutralization of the 6-month viral isolate by baseline infant plasma (i00) was 68% compared to 90% by baseline maternal plasma (m00) at the same dilution (1:20), indicating that only a subset of the maternal neutralizing antibody repertoire was passively transferred to the child. As expected, the ability of the contemporaneous plasma to neutralize the 6-month viral isolate (85%) was less than that achieved by 12, 24, 48 and 67-month plasmas, which achieved 91%, 95%, 89% and 90% neutralization, respectively (Figure 6A). The increase in neutralization by 6- to 67-month plasma as compared to at birth plasma suggested that de novo humoral immune responses against early viral genotypes persisted and became progressively stronger with time (Figure 6A). Evaluation of the contemporaneous plasma neutralization of the 12-month infant viral isolate indicated a very low level of activity (Figure 6B). Only 15% of the input virus was neutralized by the infant 12-month plasma at a 1:20 dilution; whereas, the infant plasma at birth neutralized 43%. This was 3-fold higher than the contemporaneous infant sera, but lower than the maternal plasma at delivery suggesting that most of the neutralizing antibody in the infant during the first months of life was of maternal origin. Moreover, during the first 12-month of infection, the level of neutralizing activity against the 12-month virus was observed to decrease with time indicating decay of the maternal humoral component. Thereafter, increasing titers of neutralizing antibody were detected in non-contemporaneous 24, 48, and 67-month plasma, which achieved 60, 66%, and 72% neutralization, respectively (Figure 6B). These data suggest the development of effective humoral immune responses in the infant. This increase in neutralizing humoral immunity may, in part, be responsible for observed increases in infant viral diversity during the same period. Evaluation of neutralization of the infant 48-month virus isolate revealed high titer neutralization from the maternal baseline plasma (84%), but very low level of neutralization from the infant's plasma at 24 or 48 months (Figure 6C). Nevertheless, the 67-month infant plasma neutralized 72% of the i48 m virus, suggesting a delayed but continuing infant immune response against the diversifying viral population.
Figure 6 Contemporaneous and non-contemporaneous plasma neutralization activity against infant 6-month (A), 12-month (B) and 48-month (C) viral isolates, determined in TZM-BL cells. The test plasma was diluted to 1:20, 1:100 and 1:500. Virus production in the supernatants was monitored by luciferase activity 2 days post infection. Luciferase activity in the control wells containing no plasma was defined as 100 %, and the neutralization titer of the test plasma was calculated relative to this value.
Discussion
Longitudinal evolution of HIV-1 subtype C has rarely been evaluated in infected children. The survival of infant 1157 for more than 6 years post-infection provided us with an opportunity to track genetic variation and phenotypic evolution in the viral envelope glycoprotein over that period. In addition, we were able to examine correlations between these viral properties and the humoral immune response of the child. Detection of HIV-1 sequences in PBMC collected from the child at birth indicated in utero infection. The pattern of genetic variation shown by phylogenetic analysis at baseline is compatible with an episode of selective transmission, as reported in previous studies [9-12,27,29]. In utero infection of infants has been reported to result in more rapid disease progression [30-32]; however, the extended survival of infant 1157 suggests the route of infection alone is not predictive of disease progression in subtype C infected children.
HIV-1 has replication and mutation rates that generate high numbers of progeny and significant genetic variation. The env gene has been calculated to diverge at a rate of about 1% per year [33]. The patterns of HIV-1 evolution in infected individuals, even for subtype B viruses, are ambiguous. Delwart et al. reported several-fold higher diversity at the early stage versus the late stage of infection [34]. In contrast, other studies have shown that viral sequences in env are more homogenous early in infection and diversify with disease progression and decline in CD4+ T cell counts [33,35-40]. Here we show that birth env sequences in the recipient child were highly homogenous, as indicated by env diversity, and were closely related to, but encompassed only a subset of the contemporary maternal variation. Genetic analysis at multiple timepoints showed that diversity in env as well as divergence from the initial infecting species increased with time up to 48 months. This increase in diversity and divergence correlated with parallel increases in non-synonymous changes. Whether such an increase is unique to this case needs to be further substantiated. Our findings contrast with those from studies of subtype B infected adults where, in patients infected with viruses that undergo co-receptor switching, the peak of diversity correlated with the development of CXCR4 utilization and the peak of divergence correlated with the maximal prevalence of CXCR4 utilizing species [33]. These phenomena are not relevant to 1157 since no alternative co-receptor usage was detected in either the mother or the child. Although X4-utilizing subtype C strains have been described [41-44], they are unusual, thus pointing to distinct evolutionary pressures on the various subtypes. Since subtype C infected individuals possess X4-expressing cells, it is likely that immunological and viral replicative selection in these individuals do not force or allow subtype C to efficiently utilize these targets or other constraints make such utilization significantly unfavorable.
Interestingly, we have observed the apparent reappearance of earlier lineages at the 67-month time point, and this is probably correlated to the decrease in viral genetic divergence at the same time point. Our observation would indicate that viral sequences, presumably emerging from latently infected cells, can reintroduce ancestral lineages and thus could lead to the decrease in divergence. It is tempting to speculate that such reintroduction might coincide with the waning of the immune response to these 'earlier' viruses in much the same way as antiretroviral therapy interruption often results in repopulation of the patient with drug-sensitive ancestral strains. Alternatively, the host environment may have altered in such a fashion that an ancestral variant becomes more viable due to higher replication fitness and decay of immune selection.
Our sequence analysis also revealed a substantial amount of variation (mutation, deletion and insertion) in env C3 and V4 regions in infant samples, implying that C3 or V4 domain is a likely target of immunological or replicative selective pressure during subtype C virus evolution and disease progression in children. The significance of C3 and V4 variation is currently under investigation. It is important to recognize that definitions of the constant and variable domains in Env are derived primarily from studies of subtype B viruses, and the patterns of sequence diversity in those isolates may not be reflected in other subtypes such as subtype C.
Our neutralization assays support the concept that the humoral immune response developed in parallel with the evolving HIV-1 envelope sequences and constitutes part of the selective pressure on the gene [45,46]. The persistence of high level neutralizing antibodies against early infant viral isolates indicated that the infant immune system is capable of developing and maintaining strong responses to eliminate the initially transmitted and replicating virus (Figure 6A). It has been shown that neutralization escape mutants with reduced sensitivity to autologous sera emerge rapidly in HIV-1 infected adults [46-48], but patients subsequently developed additional neutralizing antibodies to the 'escape' viruses after a delay [49]. The initial effectiveness of the infant sera is likely due to a significant contribution by maternal antibodies to neutralization titer. Nevertheless, the child does not receive the full repertoire of maternal neutralizing antibody since a disparity was observed between the effectiveness of maternal and infant baseline neutralization titers. This idea is reinforced by the fact that the maternal baseline serum continues to be effective against the infant viruses for the duration of infection; whereas the ability of the infant serum to neutralize contemporary viruses is reduced after the early timepoints. Moreover, differences in the susceptibility of viral isolates to be neutralized by antibodies was independent of the replication rates, since the 6, 12 and 48-month viral isolates replicated with nearly identical kinetics.
The observed viral diversity increase at 12-months might coincide with the diminution of maternal antibody effectiveness. However, the increasing titer of antibodies beyond 12 month implied the development of de novo infant humoral immune responses against the diversifying population. This response, as might be anticipated, is always in reaction to the viral alterations, not in anticipation of it. This conclusion is supported by the finding that despite an apparent failure of the humoral immunity to control HIV-1 replication through neutralizing antibodies at 48 months, infant 1157 mounted an effective neutralizing response to that virus at subsequent timepoints (67 month) (Figure 6C) and this coincided with a decrease in viral diversity (Figure 2). However, the role of cell-mediated immunity in controlling viral replication cannot be determined for this infant since viable cells were not available.
It has been suggested that a more antigenically diverse virus population would correlate to a broader immune reactivity, a slower rate of disease progression [50,51], and selection of neutralization escape mutants in HIV-1 infected individuals, including long-term non-progressors [47,52-54]. Our study, even though with only one mother infant pair, appears to support this hypothesis but further analysis involving a larger number of patients, including rapid and slow progressors, followed longitudinally will be needed to substantiate this observation. A more complete understanding of the mechanisms of humoral immune escape with a more precise definition of the regions in Env where such mutations cluster is likely to impact vaccine design.
It has recently been observed that viruses with shorter V1-V4 Env length, and fewer glycans are more susceptible to neutralizing antibodies, but mediate more efficient transmission in discordant couples [28]. Assuming this concept, one would expect to see a relative lengthening of Env, and an increase in the number of glycans with time. Our analysis of MIP 1157 longitudinally, which was based on C2-V4 sequences, cannot be used for direct comparison for transmission, we did, however, observe increases in variation at PNGS and in sequence length over time. The variation in Env over the follow-up period frequently resulted in the deletion, addition, or relocation of potential N-glycans, suggesting a role of N-glycans for immune selection in the HIV-1 evolution. The hot spots of N-glycan variation were particularly evident in the C2 and C3 regions. Similar changes in potential glycosylation sites have been hypothesized to modify a "glycan shield" for evading neutralizing antibodies [48].
Conclusion
We have demonstrated that genetic diversification in the infant sequences increased after 12 months, and this coincided with increases in neutralizing antibody titers. In addition, episodes of viral growth and successive immune reactions in the first 5–6 years were observed in this slow progressor infant with delayed onset of AIDS. Longitudinal studies such as the one described here underscore the dynamic and complex interactions of viral populations and immune responses. Whether this pattern of viral host interaction is typical of slow progressing infected infant needs to be further substantiated.
Methods
Patient population and sample collection
The mother-infant pair (MIP) 1157 characterized in this study was recruited to investigate the routes of transmission of HIV-1. Venous blood was obtained from the mother before delivery and from the infant within 24 hours of delivery. Follow-up blood specimens were obtained when the pair returned for visits at 6, 12, 18, 24, 29, 36, 48 and 67-months after delivery. The HIV-1 serological status of the mother was determined by two rapid assays, Capillus (Cambridge Biotech, Ireland) and Determine (Abbott laboratories, USA), on the initial blood samples. The positive serological result was confirmed by immunofluorescence assay (IFA), as previously described [55]. The status of HIV-1 infection in the infant was determined by performing viral isolation from the infant's peripheral blood mononuclear cells (PBMC) and by PCR on DNA isolated on the day of birth.
Viral isolation
HIV-1 was isolated sequentially over a 67-month post-delivery period by standard co-culture procedures. Donor HIV-1-negative PBMC were purified using Lymphoprep (Life Technology). The purified lymphocytes were then propagated in RPMI 1640 medium containing 10% heat-inactivated fetal bovine serum (FBS) and 5 μg/ml of phytohemagglutinin (Sigma) for 40 h before co-culturing with MIP 1157 PBMC or whole blood at a combined final concentration of 2 × 106cells /ml. Equal numbers of fresh uninfected PHA-stimulated PBMC were added to the culture weekly. Virus production was monitored using a commercial ELISA to measure HIV-1 p24 antigen levels (Coulter immunology, FL). Virus stocks were prepared when p24 antigen concentration exceeded 10 ng /ml (about 7–10 days). Viral isolates were recovered from 6-month maternal and 6, 12, 18, 24, 29, and 48-month infant samples
Biological phenotype
Phenotype, syncytium-inducing (SI) or non-syncytium-inducing (NSI), was determined by infecting MT-2 cells in a 12-well tissue culture plate (5 × 105 cells / well) with 5 ng of p24 virus stock per well. Cell cultures were observed daily for syncytia formation, over a course of 10 days. Levels of p24 antigen were determined in supernatants collected on day 2, 4, 7, and 10 post-infection. Virus was scored as SI if syncytia formation and increasing level of p24 antigen were observed within the 10-day period, and as NSI if syncytia failed to form within that time.
Cell tropism
To define the viral tropism, primary monocyte-derived macrophages (MDM), and MT-2 or C8166 T-cell lines were infected with the virus stocks using standard methods. Primary monocytes were obtained from gradient-purified PBMC by adherence to plastic culture dishes [56]. Adherent cells were cultured for 7 to 10 days in RPMI 1640 medium containing 10% FBS and 10 ng/ml of granulocyte-macrophage, colony-stimulating factor (GIBCO) to promote differentiation of monocytes to macrophages. Differentiated macrophages, or T-cell lines, were infected with 5 to 10 ng of HIV-1 p24 antigen per 5 × 105 cells and incubated for 4 to 5 h at 37°C. Subsequently, the infected cells were washed twice with phosphate buffered saline (PBS) and resuspended in fresh culture medium. Culture supernatants were removed at 3, 7, and 14 days post-infection and assayed for HIV-1 p24 antigen. A culture well was considered virus-positive if increasing level of p24 antigen was observed.
Chemokine co-receptor usage
Determination of co-receptor usage was carried out using cell lines obtained through the NIH AIDS Research and Reference Reagent Program, Division of AIDS, NIAID, NIH from Dr. Nathaniel Landau that express specific co-receptors (CEMx174-GFP cells [CXCR4], Ghost-CCR5 cells [CCR5] and HOS-CD4-CCR3 cells [CCR3]). PBMC from an individual homozygous for CCR5 mutation Δ32 were obtained from Dr. James Hoxie (University of Pennsylvania). To test for co-receptor usage, the CCR5-Δ32 PBMC and the three co-receptor-specific cell lines were seeded at a density of 1 × 106 cells/ml into 24-well culture plates. The cells were infected with 5 ng /ml of HIV-1 p24. The infected CEMx174-GFP and Ghost-CCR-5 cells were observed microscopically on day 2–3 post-infection for green fluorescent protein (GFP) expression. Wells exhibiting a count of GFP-expressing cells greater than or equal to 3-fold the negative control wells were scored as positive. Uninfected control wells produced only one to two GFP expressing cells per well. HIV-1 strains SF2 and NL4-3 were used as positive controls for viruses that use CXCR4, and HIV-1 strain SF128A was used as positive control for CCR5 utilization. Positive control viruses consistently gave 7-fold, or greater, GFP- expressing cells than the background control.
Infection of CCR5-Δ32 PBMC and HOS-CD4-CCR3 cells was monitored by measuring HIV-1 p24 antigen production in culture supernatants. HIV-1 p24 was measured at 3 days post-infection for HOS-CD4-CCR3 cells, and at 3, 7 and 10 days post-infection for the CCR5-Δ32 PBMC. Cultures were considered positive for viral growth if more than 100 pg/ml of p24 was detected.
Viral isolates replication kinetics
Equivalent infectious units, 100 TCID50, of the infant viral isolates obtained at 6, 12 and 48-month follow up were added to triplicate wells in a 12-well plate containing 6 × 106 PHA-stimulated PBMC from a HIV-1 seronegative blood donor. The laboratory isolate 128A was used as a replication kinetics control. After incubation at 37°C for 6 hours, cells were washed 3 times with PBS and refilled with fresh medium. All infected cultures were sampled and supplemented with a 50% volume of fresh culture medium at day 1, 3, 5, 7, 9, 11, 13 and 15. Viral replication kinetics in PBMC was determined by measuring RT units in culture supernatants at day 0, 1, 3, 5, 7, 9, 11, 13 and 15-postinfection. Viruses were lysed using 10% Triton X-100 (1% final concentration) in RPMI medium supplemented with 10% FBS, and RT units was measured using EnzChek Reverse Transcriptase Assay Kit (Invitrogen, Eugene, Oregon). The assay was performed in triplicate.
Virus neutralization assay
Plasma neutralization activity was determined through infections of TZM-bl cells (NIH AIDS Research and Reference Reagent Program catalogy no. 8129, TZM-bl) as described in Wei et al (2003) with modifications. TZM-bl cells stably express high levels of CD4, CCR5 and CXCR4. The cells contain HIV-1 LTR promoter cassettes that express luciferase and β-galactosidase in response to stimulation with HIV-1 Tat. TZM-bl cells were plated at a density of 6 × 103/well in 96-well tissue culture plates (Falcon) and cultured overnight in DMEM supplemented with 10% FBS. Test plasma was heat-inactivated at 56°C for 30 min, spun at 3,000 × g for 5 min and diluted 1:20, 1:100 and 1:500 in DMEM plus 6% FBS. Viral aliquots of 100 TCID50/ml were prepared in DMEM supplemented with 6% FBS and 80 μg/ml DEAE dextran, to a combined total volume of 100 μl. The virus aliquots (100 μl) were combined with 100 μl of the different test plasma dilutions and the mixture was incubated for 1h at 37°C. Following incubation, the virus-plasma mixture was added to TZM- bl cells and incubated at 37°C for two days. Following two washes with PBS, the level of virus infection was measured by luciferase activity. Cells were lysed using Luciferase Assay Reagent (Promega, Madison, WI) and the luciferase activity was measured using a LUMIstar luminometer (BMG Lab Technologies, Offenburg, Germany). The assay was performed in triplicate. Controls included cells infected by virus inoculated with medium or normal human plasma instead of the test plasma. Effective neutralization by the plasma would reduce the level of luciferase versus controls lacking plasma. The luciferase activity in the control wells without plasma was defined as 100 %, and the neutralization titer of the test plasma was calculated relative to this value.
Polymerase chain reaction, gene cloning, sequencing and subtype identification
Genomic DNA was purified from patient PBMC using the ISOQUICK kit (ORCA Research, Inc.). Primers used for amplification of the subtype C env gene were designed based on a reference alignment of all HIV-1 subtypes obtained from the Los Alamos HIV-1 Sequence Database . Nested PCR, as previously described [27], was used to amplify a 617 bp fragment containing the C2-V4 region of the maternal and infant env gene from samples collected at birth through 67-month. The amplified fragments were cloned into the pGEM-T Easy vector (Promega) and sequenced in both directions with dideoxy terminators (ABI BigDye Kit).
Sequence alignment and analyses
Nucleotide sequences were translated and the amino acid sequences aligned using Clustal W, as implemented in BioEdit (Hall 1999)[57], and further refined manually, preserving the reading frame.
Bayesian (BA), maximum likelihood (ML), and neighbor-joining (NJ) phylogenetic analyses were implemented to: 1) determine subtype affiliation (aligning infant clones to a reference panel from the HIV database at Los Alamos National Laboratory ; 2) assess the transmission pattern between 1157 mother and infant (samples collected at birth from the mother and the infant were evaluated), 3) visualize how viral populations change with time (all samples collected from mother and infant were included). Bayesian searches were run in MrBayes version 3.04 [58] for 5 × 106 generations. Trees were sampled every 1000 generations. Convergence was reached after 1.5 × 106, and a majority rule consensus was built based on the last 3000 trees. ML searches were implemented in Treefinder [59] using a general time-reversible model of nucleotide substitution for each codon position. NJ analyses were performed in PAUP* ver4.10b [60], estimating distances by maximum likelihood, after selecting the best fitting model of nucleotide substitution in Modeltest ver 3.5 [61]. Support for the nodes in ML and NJ was evaluated by running 1000 pseudoreplicates in bootstrap analyses.
Genetic diversity, temporal structure and estimates of selective strength
Samples were grouped into contemporaneous sets according to the time of collection. Sites with alignment gaps, corresponding to putative insertions or deletions, were excluded from comparisons. Variation in the pattern of genetic diversity for each time-point was explored using MEGA3, [62] and DNAsp ver 4.06 [63], Datamonkey [64] and PAML ver 3.13. [65], and the number and location of putative glycosylation sites (PNGS) were estimated using N-GlycoSite from the Los Alamos National Laboratory database . Viral diversity and viral divergence from the initial population were calculated for each timepoint. Viral diversity estimates were based on π, the average nucleotide differences between sequences per site. Changes in viral diversity were used to estimate relative changes in the effective viral population size, assuming a similar mutation rate at each timepoint. Viral divergence was calculated as the average genetic distance to the viral population in the infant at birth.
The relative strength of positive selection was evaluated in a ML framework using PAML ver 3.13. [65]. To estimate the overall dN/dS for each set of contemporaneous sequences we implemented the one-rate codon model described by Yang and Nielsen (1998) [66]. The number of synonymous and non-synonymous substitutions observed per site in the infant Env domains (C2-V4) was calculated in Datamonkey so as to visualize the distribution of synonymous and non-synonymous substitutions along the sequence and to identify regions that accumulate most variation ('hot spots').
Authors' contributions
HZ carried out the PCR, cloning, and sequencing. FH, GO, HZ and XH performed the sequencing analysis by computer program. JH carried out viral isolation, viral tropism, co-receptor usage and neutralization assay. CK was involved in patient recruitment and follow-up. RR contributed to experimental design and data analysis. HZ, FH, GO, JW and CW participated in the experimental design, data interpretation and writing of the manuscript.
Acknowledgements
This study was supported by PHS grants HD39620, CA76958, PO1 AI 48240, Fogarty Training grant TW01429, and NCRR COBRE grant RR15635 to C.W, PO1 AI34266 to R.M.R. We thank Dr. James Hoxie and for providing us with CCR5 Δ 32 cells; Dr. Nathaniel Landau for HOS-CD4-CCR3 cells through the NIH AIDS research and reference reagent Program, Division of AIDS.
==== Refs
Essex M State of the HIV pandemic J Hum Virol 1998 1 427 429 10195263
Janssens W Buve A Nkengasong JN The puzzle of HIV-1 subtypes in Africa AIDS 1997 11 705 712 9143601 10.1097/00002030-199706000-00002
Rodenburg CM Li Y Trask SA Chen Y Decker J Robertson DL Kalish ML Shaw GM Allen S Hahn BH Gao F Near full-length clones and reference sequences for subtype C isolates of HIV type 1 from three different continents AIDS Res Hum Retroviruses 2001 17 161 168 11177395 10.1089/08892220150217247
Luzuriaga K Sullivan JL Pediatric HIV-1 infection: advances and remaining challenges AIDS Rev 2002 4 21 26 11998780
Edgeworth RL Ugen KE Immunopathological factors for vertical transmission of HIV-1 Pathobiology 2000 68 53 67 10878502 10.1159/000028115
Arroyo MA Tien H Pagan M Swanstrom R Hillyer GV Cadilla CL Melendez-Guerrero LM Virologic risk factors for vertical transmission of HIV type 1 in Puerto Rico AIDS Res Hum Retroviruses 2002 18 447 460 11958688 10.1089/088922202753614218
Garcia PM Kalish LA Pitt J Minkoff H Quinn TC Burchett SK Kornegay J Jackson B Moye J Hanson C Zorrilla C Lew JF Maternal levels of plasma human immunodeficiency virus type 1 RNA and the risk of perinatal transmission. Women and Infants Transmission Study Group N Engl J Med 1999 341 394 402 10432324 10.1056/NEJM199908053410602
John GC Nduati RW Mbori-Ngacha DA Richardson BA Panteleeff D Mwatha A Overbaugh J Bwayo J Ndinya-Achola JO Kreiss JK Correlates of mother-to-child human immunodeficiency virus type 1 (HIV-1) transmission: association with maternal plasma HIV-1 RNA load, genital HIV-1 DNA shedding, and breast infections J Infect Dis 2001 183 206 212 11120927 10.1086/317918
Dickover RE Garratty EM Plaeger S Bryson YJ Perinatal transmission of major, minor, and multiple maternal human immunodeficiency virus type 1 variants in utero and intrapartum J Virol 2001 75 2194 2203 11160723 10.1128/JVI.75.5.2194-2203.2001
Wolinsky SM Wike CM Korber BT Hutto C Parks WP Rosenblum LL Kunstman KJ Furtado MR Munoz JL Selective transmission of human immunodeficiency virus type-1 variants from mothers to infants Science 1992 255 1134 1137 1546316
Scarlatti G Leitner T Halapi E Wahlberg J Marchisio P Clerici-Schoeller MA Wigzell H Fenyo EM Albert J Uhlen M Rossi P Comparison of variable region 3 sequences of human immunodeficiency virus type 1 from infected children with the RNA and DNA sequences of the virus populations of their mothers Proc Natl Acad Sci USA 1993 90 1721 1725 8446584
Ahmad N Baroudy BM Baker RC Chappey C Genetic analysis of human immunodeficiency virus type 1 envelope V3 region isolates from mothers and infants after perinatal transmission J Virol 1995 69 1001 1012 7815476
Pasquier C Cayrou C Blancher A Tourne-Petheil C Berrebi A Tricoire J Puel J Izopet J Molecular evidence for mother-to-child transmission of multiple variants by analysis of RNA and DNA sequences of human immunodeficiency virus type 1 J Virol 1998 72 8493 8501 9765386
Lamers SL Sleasman JW She JX Barrie KA Pomeroy SM Barrett DJ Goodenow MM Persistence of multiple maternal genotypes of human immunodeficiency virus type I in infants infected by vertical transmission J Clin Invest 1994 93 380 390 8282808
Chakraborty R HIV-1 infection in children: a clinical and immunologic overview Curr HIV Res 2005 3 31 41 15638721 10.2174/1570162052773022
Hwang SS Boyle TJ Lyerly HK Cullen BR Identification of the envelope V3 loop as the primary determinant of cell tropism in HIV-1 Science 1991 253 71 74 1905842
Fouchier RA Groenink M Kootstra NA Tersmette M Huisman HG Miedema F Schuitemaker H Phenotype-associated sequence variation in the third variable domain of the human immunodeficiency virus type 1 gp120 molecule J Virol 1992 66 3183 3187 1560543
de Jong JJ Goudsmit J Keulen W Klaver B Krone W Tersmette M de Ronde A Human immunodeficiency virus type 1 clones chimeric for the envelope V3 domain differ in syncytium formation and replication capacity J Virol 1992 66 757 765 1731110
Milich L Margolin B Swanstrom R V3 loop of the human immunodeficiency virus type 1 Env protein: interpreting sequence variability J Virol 1993 67 5623 5634 8350415
Stamatatos L Cheng-Mayer C Evidence that the structural conformation of envelope gp120 affects human immunodeficiency virus type 1 infectivity, host range, and syncytium-forming ability J Virol 1993 67 5635 5639 8350416
Richman DD Bozzette SA The impact of the syncytium-inducing phenotype of human immunodeficiency virus on disease progression J Infect Dis 1994 169 968 974 7909549
Cocchi F DeVico AL Garzino-Demo A Cara A Gallo RC Lusso P The V3 domain of the HIV-1 gp120 envelope glycoprotein is critical for chemokine-mediated blockade of infection Nat Med 1996 2 1244 1247 8898753 10.1038/nm1196-1244
Connor RI Sheridan KE Ceradini D Choe S Landau NR Change in coreceptor use coreceptor use correlates with disease progression in HIV-1 – infected individuals J Exp Med 1997 185 621 628 9034141 10.1084/jem.185.4.621
Hayman A Moss T Arnold C Naylor-Adamson L Balfe P Disease progression in heterosexual patients infected with closely related subtype B strains of HIV type 1 with differing coreceptor usage properties AIDS Res Hum Retroviruses 2004 20 365 371 15157355 10.1089/088922204323048113
Nabatov AA Pollakis G Linnemann T Kliphius A Chalaby MI Paxton WA Intrapatient alterations in the human immunodeficiency virus type 1 gp120 V1V2 and V3 regions differentially modulate coreceptor usage, virus inhibition by CC/CXC chemokines, soluble CD4, and the b12 and 2G12 monoclonal antibodies J Virol 2004 78 524 530 14671134 10.1128/JVI.78.1.524-530.2004
Oelrichs RB Shrestha IL Anderson DA Deacon NJ The explosive human immunodeficiency virus type 1 epidemic among injecting drug users of Kathmandu, Nepal, is caused by a subtype C virus of restricted genetic diversity J Virol 2000 74 1149 1157 10627525 10.1128/JVI.74.3.1149-1157.2000
Zhang H Orti G Du Q He J Kankasa C Bhat G Wood C Phylogenetic and phenotypic analysis of HIV type 1 env gp120 in cases of subtype C mother-to-child transmission AIDS Res Hum Retroviruses 2002 18 1415 1423 12512513 10.1089/088922202320935492
Derdeyn CA Decker JM Bibollet-Ruche F Mokili JL Muldoon M Denham SA Heil ML Kasolo F Musonda R Hahn BH Shaw GM Korber BT Allen S Hunter E Envelope-constrained neutralization-sensitive HIV-1 after heterosexual transmission Science 2004 303 2019 2022 15044802 10.1126/science.1093137
Verhofstede C Demecheleer E De Cabooter N Gaillard P Mwanyumba F Claeys P Chohan V Mandaliya K Temmerman M Plum J Diversity of the human immunodeficiency virus type 1 (HIV-1) env sequence after vertical transmission in mother-child pairs infected with HIV-1 subtype A J Virol 2003 77 3050 3057 12584330 10.1128/JVI.77.5.3050-3057.2003
Blanche S Rouzioux C Moscato ML Veber F Mayaux MJ Jacomet C Tricoire J Deville A Vial M Firtion G Crepy AD Douard D Robin M Courptin C Ciraru-Vigneron N Diest F Griscelli C A prospective study of infants born to women seropositive for human immunodeficiency virus type 1. HIV Infection in Newborns French Collaborative Study Group N Engl J Med 1989 320 1643 1648 2657430
Blanche S Tardieu M Duliege A Rouzioux C Le Deist F Fukunaga K Caniglia M Jacomet C Messiah A Griscelli C Longitudinal study of 94 symptomatic infants with perinatally acquired human immunodeficiency virus infection. Evidence for a bimodal expression of clinical and biological symptoms Am J Dis Child 1990 144 1210 1215 1978551
Mayaux MJ Burgard M Teglas JP Cottalorda J Krivine A Simon F Puel J Tamalet C Dormont D Masquelier B Doussin A Rouzioux C Blanche S Neonatal characteristics in rapidly progressive perinatally acquired HIV-1 disease. The French Pediatric HIV Infection Study Group Jama 1996 275 606 610 8594241 10.1001/jama.275.8.606
Shankarappa R Margolick JB Gange SJ Rodrigo AG Upchurch D Farzadegan H Gupta P Rinaldo CR Learn GH He X Huang XL Mullins JI Consistent viral evolutionary changes associated with the progression of human immunodeficiency virus type 1 infection J Virol 1999 73 10489 10502 10559367
Delwart EL Pan H Sheppard HW Wolpert D Neumann AU Korber B Mullins JI Slower evolution of human immunodeficiency virus type 1 quasispecies during progression to AIDS J Virol 1997 71 7498 7508 9311829
Hutto C Zhou Y He J Geffin R Hill M Scott W Wood C Longitudinal studies of viral sequence, viral phenotype, and immunologic parameters of human immunodeficiency virus type 1 infection in perinatally infected twins with discordant disease courses J Virol 1996 70 3589 3598 8648693
Markham RB Wang WC Weisstein AE Wang Z Munoz A Templeton A Margolick J Vlahov D Quinn T Farzadegan H Yu XF Patterns of HIV-1 evolution in individuals with differing rates of CD4 T cell decline Proc Natl Acad Sci USA 1998 95 12568 12573 9770526 10.1073/pnas.95.21.12568
Nowak MA Bangham CR Population dynamics of immune responses to persistent viruses Science 1996 272 74 79 8600540
Salvatori F Masiero S Giaquinto C Wade CM Brown AJ Chieco-Bianchi L De Rossi A Evolution of human immunodeficiency virus type 1 in perinatally infected infants with rapid and slow progression to disease J Virol 1997 71 4694 4706 9151863
Joos B Trkola A Fischer M Kuster H Rusert P Leemann C Boni J Oxenius A Price DA Phillips RE Wong JK Hirschel B Weber R Gunthard HF Low human immunodeficiency virus envelope diversity correlates with low in vitro replication capacity and predicts spontaneous control of plasma viremia after treatment interruptions J Virol 2005 79 9026 9037 15994796 10.1128/JVI.79.14.9026-9037.2005
Troyer RM Collins KR Abraha A Fraundorf E Moore DM Krizan RW Toossi Z Colebunders RL Jensen MA Mullins JI Vanham G Arts EJ Changes in human immunodeficiency virus type 1 fitness and genetic diversity during disease progression J Virol 2005 79 9006 9018 15994794 10.1128/JVI.79.14.9006-9018.2005
Abebe A Demissie D Goudsmit J Brouwer M Kuiken CL Pollakis G Schuitemaker H Fontanet AL Rinke de Wit TF HIV-1 subtype C syncytium- and non-syncytium-inducing phenotypes and coreceptor usage among Ethiopian patients with AIDS AIDS 1999 13 1305 1311 10449282 10.1097/00002030-199907300-00006
Cilliers T Nhlapo J Coetzer M Orlovic D Ketas T Olson WC Moore JP Trkola A Morris L The CCR5 and CXCR4 Coreceptors Are Both Used by Human Immunodeficiency Virus Type 1 Primary Isolates from Subtype C J Virol 2003 77 4449 4456 12634405 10.1128/JVI.77.7.4449-4456.2003
Johnston ER Zijenah LS Mutetwa S Kantor R Kittinunvorakoon C Katzenstein DA High Frequency of Syncytium-Inducing and CXCR4-Tropic Viruses among Human Immunodeficiency Virus Type 1 Subtype C-Infected Patients Receiving Antiretroviral Treatment J Virol 2003 77 7682 7688 12805470 10.1128/JVI.77.13.7682-7688.2003
Pollakis G Abebe A Kliphuis A Chalaby MIM Bakker M Mengistu Y Brouwer M Goudsmit J Schuitemaker H Paxton WA Phenotypic and Genotypic Comparisons of CCR5- and CXCR4-Tropic Human Immunodeficiency Virus Type 1 Biological Clones Isolated from Subtype C-Infected Individuals J Virol 2004 78 2841 2852 14990703 10.1128/JVI.78.6.2841-2852.2004
Korber B Gaschen B Yusim K Thakallapally R Kesmir C Detours V Evolutionary and immunological implications of contemporary HIV-1 variation Br Med Bull 2001 58 19 42 11714622 10.1093/bmb/58.1.19
Richman DD Wrin T Little SJ Petropoulos CJ Rapid evolution of the neutralizing antibody response to HIV type 1 infection Proc Natl Acad Sci USA 2003 100 4144 4149 12644702 10.1073/pnas.0630530100
Arendrup M Nielsen C Hansen JE Pedersen C Mathiesen L Nielsen JO Autologous HIV-1 neutralizing antibodies: emergence of neutralization-resistant escape virus and subsequent development of escape virus neutralizing antibodies J Acquir Immune Defic Syndr 1992 5 303 307 1740756
Wei X Decker JM Wang S Hui H Kappes JC Wu X Salazar-Gonzalez JF Salazar MG Kilby JM Saag MS Komarova NL Nowak MA Hahn BH Kwong PD Shaw GM Antibody neutralization and escape by HIV-1 Nature 2003 422 307 312 12646921 10.1038/nature01470
Khouri YF McIntosh K Cavacini L Posner M Pagano M Tuomala R Marasco WA Vertical Transmission of HIV-1. Correlation with maternal viral load and plasma levels of CD4 binding site anti-gp120 antibodies J Clin Invest 1995 95 732 737 7860754
Ganeshan S Dickover RE Korber BT Bryson YJ Wolinsky SM Human immunodeficiency virus type 1 genetic evolution in children with different rates of development of disease J Virol 1997 71 663 677 8985398
Halapi E Leitner T Jansson M Scarlatti G Orlandi P Plebani A Romiti L Albert J Wigzell H Rossi P Correlation between HIV sequence evolution, specific immune response and clinical outcome in vertically infected infants Aids 1997 11 1709 1717 9386805 10.1097/00002030-199714000-00007
Bradney AP Scheer S Crawford JM Buchbinder SP Montefiori DC Neutralization escape in human immunodeficiency virus type 1-infected long-term nonprogressors J Infect Dis 1999 179 1264 1267 10191234 10.1086/314711
Lathey JL Pratt RD Spector SA Appearance of autologous neutralizing antibody correlates with reduction in virus load and phenotype switch during primary infection with human immunodeficiency virus type 1 J Infect Dis 1997 175 231 232 8985228
McGrath KM Hoffman NG Resch W Nelson JA Swanstrom R Using HIV-1 sequence variability to explore virus biology Virus Res 2001 76 137 160 11410314 10.1016/S0168-1702(01)00271-4
Mantina H Kankasa C Klaskala W Brayfield B Campbell J Du Q Bhat G Kasolo F Mitchell C Wood C Vertical transmission of Kaposi's sarcoma-associated herpesvirus Int J Cancer 2001 94 749 752 11745472 10.1002/ijc.1529
Cheng-Mayer C Shioda T Levy JA Host range, replicative, and cytopathic properties of human immunodeficiency virus type 1 are determined by very few amino acid changes in tat and gp120 J Virol 1991 65 6931 6941 1658383
Thompson JD Gibson TJ Plewniak F Jeanmougin F Higgins DG The CLUSTAL_X windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools Nucleic Acids Res 1997 25 4876 4882 9396791 10.1093/nar/25.24.4876
Ronquist F Huelsenbeck JP MrBayes 3: Bayesian phylogenetic inference under mixed models Bioinformatics 2003 19 1572 1574 12912839 10.1093/bioinformatics/btg180
Jobb G von Haeseler A Strimmer K TREEFINDER: a powerful graphical analysis environment for molecular phylogenetics BMC Evol Biol 2004 4 18 15222900 10.1186/1471-2148-4-18
Swofford DL PAUP* Phylogenetic Analysis using parsimony and other methods Version 40 beta edn 1998 Sunderland, MA: Sinauer Associates
Posada D Crandall KA MODELTEST: testing the model of DNA substitution Bioinformatics 1998 14 817 818 9918953 10.1093/bioinformatics/14.9.817
Kumar S Tamura K Nei M MEGA: Molecular Evolutionary Genetics Analysis 101 edn 1993 University Park, PA 16802: The Pennsylvania State University
Rozas J Sanchez-DelBarrio JC Messeguer X Rozas R DnaSP, DNA polymorphism analyses by the coalescent and other methods Bioinformatics 2003 19 2496 2497 14668244 10.1093/bioinformatics/btg359
Kosakovsky Pond SL Frost SDW Datamonkey: rapid detection of selective pressure on individual sites of codon alignments Bioinformatics 2005 bti320
Yang Z PAML: a program package for phylogenetic analysis by maximum likelihood Comput Appl Biosci 1997 13 555 556 9367129
Nielsen R Yang Z Likelihood models for detecting positively selected amino acid sites and applications to the HIV-1 envelope gene Genetics 1998 148 929 936 9539414
|
16274482
|
PMC1308862
|
CC BY
|
2021-01-04 16:36:41
|
no
|
Retrovirology. 2005 Nov 7; 2:67
|
utf-8
|
Retrovirology
| 2,005 |
10.1186/1742-4690-2-67
|
oa_comm
|
==== Front
RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-681627448410.1186/1742-4690-2-68ResearchThe pp24 phosphoprotein of Mason-Pfizer monkey virus contributes to viral genome packaging Bohl Christopher R [email protected] Shanna M [email protected] Robert A [email protected] School of Biological Sciences and the Nebraska Center for Virology, University of Nebraska, Lincoln, 68588, USA2005 7 11 2005 2 68 68 12 4 2005 7 11 2005 Copyright © 2005 Bohl et al; licensee BioMed Central Ltd.2005Bohl et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The Gag protein of Mason-Pfizer monkey virus, a betaretrovirus, contains a phosphoprotein that is cleaved into the Np24 protein and the phosphoprotein pp16/18 during virus maturation. Previous studies by Yasuda and Hunter (J. Virology. 1998. 72:4095–4103) have demonstrated that pp16/18 contains a viral late domain required for budding and that the Np24 protein plays a role during the virus life cycle since deletion of this N-terminal domain blocked virus replication. The function of the Np24 domain, however, is not known.
Results
Here we identify a region of basic residues (KKPKR) within the Np24 domain that is highly conserved among the phosphoproteins of various betaretroviruses. We show that this KKPKR motif is required for virus replication yet dispensable for procapsid assembly, membrane targeting, budding and release, particle maturation, or viral glycoprotein packaging. Additional experiments indicated that deletion of this motif reduced viral RNA packaging 6–8 fold and affected the transient association of Gag with nuclear pores.
Conclusion
These results demonstrate that the Np24 domain plays an important role in RNA packaging and is in agreement with evidence that suggests that correct intracellular targeting of Gag to the nuclear compartment is an fundamental step in the retroviral life cycle.
==== Body
Introduction
Viruses of the Betaretroviruses genus, formerly known as D- and B-type retroviruses, assemble their capsids in the cytoplasm of infected cells instead of at the plasma membrane like most retroviruses. The B-type viruses contain prominent surface glycoproteins and spherical, eccentric capsids and include mouse mammary tumor virus (MMTV) and exogenous and endogenous MMTV-like retroviruses in mice and humans [1-3]. D-type viruses have less dense surface spikes and contain cylindrical capsids. Exogenous and endogenous D-type viruses infect in a variety of mammalian hosts including Old World monkeys (Mason-Pfizer monkey virus [M-PMV], simian retrovirus 1 [SRV-1], [SRV-2] and simian endogenous retrovirus) [4-6], New World monkeys (squirrel monkey retrovirus [SMRV]) [7], sheep and goats (Jaagsiekte sheep retrovirus and enzootic nasal tumor virus respectively) [8-10]. D-type virus sequences have also been detected in humans, the Australian common brushtail possum and mice (Trichosurus vulpecula endogenous retrovirus D, rabbit endogenous virus H, and MusD, respectively) [11-13].
M-PMV, the prototypical D-type virus, was first isolated from a mammary adenocarcinoma of a female Rhesus monkey [14]. Although M-PMV was originally suspected to be an oncogenic virus, it was later found to induce a sever "wasting" and immunodeficiency syndrome distinct from that caused by immunosuppressive lentiviruses [15]. SRV-1 and SRV-2 are related to, yet serotypically distinct from, M-PMV and were isolated from primates suffering diseases similar to that caused by M-PMV [16,17].
M-PMV, the most thoroughly understood of the D-type betaretroviruses, contains four genes (5'-gag-pro-pol-env). As with other retroviruses, its Gag protein, Pr78, serves multiple functions during the viral life cycle, including virus assembly, virion maturation and early post-entry steps in virus replication [18]. Multiple studies have shown that Pr78 has the innate ability to assemble into immature capsids or procapsids in the cytoplasm, recognize and package the viral RNAs and glycoproteins and facilitate budding from the plasma membrane. During viral budding or shortly thereafter, Pr78 is cleaved by the viral protease to yield the mature virion associated proteins: matrix MA (p10), the phosphoprotein pp24, p12, capsid (CA or p27), nucleocapsid (NC or p14) and p4. These mature Gag-cleavage products then play roles during the early stages of the viral life cycle where they may help facilitate uncoating, reverse transcription and nuclear entry of the viral DNA. The regions and modifications of Pr78 required for these events have been partially identified.
Upon translation, Pr78 is targeted to a pericentriolar region of the cytoplasm in close proximity to the nuclear membrane where it assembles into spherical, procapsids [19]. The signal within Pr78 responsible for this pericentriolar targeting (the cytoplasmic targeting/retention signal or CTRS) is located within an 18 amino acid sequence of the matrix domain (MA). This motif is dominant over the bipartite myristylation and lysine/arginine-rich bipartite membrane targeting signals that is also located within the MA domain. Insertion of the CTRS into the analogous region of the MLV Gag protein, which normally assembles at the plasma membrane, results in intracytoplasmic assembly of MLV Gag. Secondly, substitution of an arginine within the CTRS of M-PMV Gag to a tryptophan (R55W) destroys the dominant CTRS function resulting in capsid assembly at the plasma membrane [20].
Other regions of Pr78 are also essential for procapsid assembly. Residues within the MA, yet separate from the CTRS, and the CA domains are required for assembly [20-23]. Likewise, the p12 domain with in Pr78 provides an internal scaffolding that together with the cononical I domain, which is located near the CA-NC junction, function to promote Gag-Gag interactions during capsid [24,25]. Assembly of the spherical capsid also requires interactions between the viral RNAs (vRNA) and Gag proteins [26,27]. Thus, the vRNA must also be present at the assembly site to provide this additional nucleation or scaffolding function. Although Gag-vRNA interaction occurs primarily though interaction in the NC domain, other regions of Gag appear to be important by targeting Gag to the site of vRNA packaging and by imparting correct structural information upon Gag [22,28-32].
Following assembly, the procapsids are transported to the plasma membrane from which they bud. Both the myristyic acid modification of Pr78 and specific amino acids within the MA domain play critical roles in plasma membrane targeting [20,21]. Moreover, Sfakianos and Hunter have shown that the M-PMV Env glycoproteins and Rab11-positive recycling endosomes play critical roles in transporting the preassembled procapsids from the pericentriolar assembly site to the plasma membrane [33]. Upon arrival of the procapsids at the plasma membrane, a proline-rich PPPYX4PSAP motif located near the carboxy-terminus of the Gag phosphoprotein, pp24, provides the late budding domain (L), which facilitates viral budding [34,35].
Upon release, nascent immature particles undergo a maturation process to acquire infectivity. During this process, Pr78 is cleaved by the viral protease to yield the seven proteins: MA, pp24, p12, CA, NC, and p4. The phosphoprotein conserved among M-PMV, SRV-1, SRV-2, SERV, and MMTV yet its function is only partially known. During M-PMV maturation, pp24 is further cleaved into two proteins; the C-terminal pp16 protein and the N-terminal Np24 protein. Both are required for virus replication. Yasuda and Hunter demonstrated that the pp16 domain contains the late budding motif and deletion of the Np24 domain completely blocked virus replication [35]. However, the function of Np24 was not determined. In this study, we examined the role of the Np24 domain during virus replication. We have identified a lysine-arginine rich motif within Np24 that is conserved among many betaretrovirus and is essential for infectivity. The results presented here show that the KKPKR motif in Np24 is not required for procapsid assembly, intracellular transport, budding or glycoprotein incorporation but plays a critical role in vRNA packaging.
Results
Deletion of the KR box in Np24 blocks virus replication
While it was previously determined that the Np24 domain of Pr78Gag is required for replication [35], the role of this protein plays during the virus life cycle is not known. To gain further insight into which regions(s) of Np24 might be important for replication, the Np24 protein sequence was aligned to the analogous phosphoproteins from different infectious betaretroviruses (Fig. 1A). As expected because of the close similarity between the simian betaretroviruses M-PMV and SRV-1 and the more distantly related SRV-2 and SERV, their phosphoprotein sequences are 82%, 62%, and 61% (respectively) similar to M-PMV Np24. The most notable similarities occur at the amino- and carboxy-terminal ends of theses phosphoproteins. While the amino-terminal sequences are not conserved in the phosphoproteins of MMTV and MMTV related betaretroviruses, the highly conserved cluster of positively charged amino acids, KKPKR, (the KR box) near the carboxy-terminal end is shared by these betaretroviruses. In M-PMV (and SRV-1), the KR box is located near the carboxy-terminal end of Np24. In MMTV, it is located in the same relative position of pp21. The conservation of the KR box within the phosphoproteins of these divergent viruses suggests that it may be essential to virus replication. To determine if this motif (KKPKR) in M-PMV serves an important role during the virus life cycle, PCR mutagenesis was used to delete the region encoding the KKPKR motif from the infectious clone pSARM4. The mutant, pΔKKPKR, and wild-type pSARM4 proviral DNAs were transfected into COS-1 cells. At 48 h post transfections the viruses produced from the transfected cells were harvested and assayed for RT activity. Wild-type M-PMV and ΔKKPKR-transfected cells produced equivalent amounts of virus particles (data not shown). Viral spread assays were then carried out to examine if the deletion of the KKPKR motif affected viral replication. For this, Hos cells were infected with equivalent amounts of wild-type and mutant virus particles, normalized by RT activities. The amounts of virus particles present in the supernatants of infected cells at 2, 4, 8, 10, and 12 days post-infection were determined by RT assays. While wild-type virus replicated in Hos cells, as indicated by the increasing amounts of RT activity in the culture medium over time, no detectable RT activity was observed in the supernatants of uninfected or ΔKKPKR-infected COS-1 cells even at 14 days post-infection (Fig. 1B). These data demonstrates that deletion of the KKPKR motif blocked virus replication.
Figure 1 The highly conserved KR box within the Np24 protein of M-PMV is required for viral replication. (A) Amino acid alignment of phosphoproteins of betaretroviruses showing areas of sequence conservation in the N-terminus and C-terminus. All conserved residues are bolded and the highly conserved KR box is shadowed. Accession Numbers; M-PMV (P07567), SRV-1 (AAA47730), SRV-2 (P51516), MMTV (AAF3147) (B) Viral replication measured by virus spread assay. Culture media from COS-1 cells transfected with nothing (Black), wild-type M-PMV proviral DNA, pSARM4 (Red), or ΔKKPKR proviral DNA (Blue) were collected and assayed for RT activity. HOS cells were infected with equal amounts of virus (normalized by RT activity). Virus spread in HOS cells was measured by RT assays at 2, 4, 6, 8, 10, 12, and 14 days post infection.
Assembly and release of ΔKKPKR mutant particles
Because the different morphogenic steps (procapsid assembly, cytoplasmic transport, membrane biding, and budding) are temporally separate for M-PMV, this virus provides an ideal opportunity to determine which if any of the late assembly steps might be affected by the deletion mutation. To determine if the replication-defective ΔKKPKR mutant could assemble intracellular procapsids, transfected cells were lysed in a TX-100 lysis buffer that does not disrupt assembled procapsids. The assembled procapsids were separated from the soluble unassembled Gag proteins in the cellular lysates by centrifugation through a 20% sucrose cushion. The pelleted proteins were solublized directly in protein loading buffer, separated by SDS-PAGE, and immunoblotted using anti-Pr78 antibodies. As expected, wild-type Gag (Pr78WT) was detected in both the soluble fraction (unassembled) and pelleted fractions (assembled procapsids) of the cellular lysates (Fig 2A. Lanes 1 and 2). The presence of the ΔKKPKR mutant Gag protein (Pr78ΔKR) in the soluble and pelletable fractions (Fig. 2A lanes 4 and 5) indicates that the mutant Gag proteins assembled into procapsids.
Figure 2 (A) Western Blot analysis of intracellular procapsid assembly using Gag fractionation techniques. 48 hrs post transfection, COS-1 cells were lysed with and fractionated over a 20% sucrose cushion to separate assembled procapsids from unassembled Gag proteins. Pr78 in the fractionated samples were detected by western blot using rabbit anti-Pr78 antibodies. Soluble wild-type Pr78 (lane 1); pelletable wild-type Pr78 (lane 2); untransfected (lane 3); soluble ΔKKPKR Pr78 (lane 4); pelletable ΔKKPKR Pr78 (lane 5). (B) Virus release kinetics. Transfected COS-1 cells pulsed labeled with [35S] methionine-cysteine for 30 minutes and chased for 0, 1, 2, 4, and 8 hours. Untransfected (lane 1); pSARM-4 (lanes 2–6); ΔKKRKR (lanes 7–11). Medium was collected and cells were lysed at the appropriate times with 1× Buffer A. Cellular lysates and medium were adjusted to 1× lysis buffer B. Viral proteins were immunoprecipitated from all samples using rabbit anti-Pr78 antibodies and separated by SDS-PAGE (12% acrylamyde) and detected by phospor-imaging.
The rate of assembly and release of capsids from Pr78WT and Pr78ΔKR expressing cells were analyzed by pulse-chase analyses to determine if the deletion had caused defects in virus assembly and release. Transfected COS-1 cells were pulse labeled with [35S] methionine-cysteine for 30 min. and chased for 1, 2, 4, and 8 hours in complete growth medium. Cell associated and released-virus-associated proteins at each time point were analyzed by immunoprecipitation using rabbit anti-Pr78 antiserum (Fig. 2)
Similar levels of Gag (Pr78), Gag-Pro (Pr95), and Gag-Pro-Pol (Pr180) fusion proteins were synthesized during the pulse labeling by cells expressing wild-type M-PMV (Fig. 2B). The 68 kDa protein is a N-terminal truncated Gag protein that is expressed by initiation of translation at an internal methionine codon of gag at position 100 [24]. As expected, the intensities of these wild-type Gag proteins in the cell lysates decreased during the chase with a concomitant appearance of p27 (CA). During or shortly after virus release, Pr78WT is processed by the virus-encoded protease into p10 (MA), Np24, pp16/18, p12, p27 (CA), p14 (NC), and p4. Thus the appearance of p27 in the culture medium indicates that virus particles were released and that Pr78WT was being processed normally (Fig. 2B, lane 4). The other Gag cleavage products were not detected because they either do not contain methionines, or contain only a single methionine and thus were not detected.
In cells expressing the mutant virus, similar levels of cell-associated Gag precursors were observed in the pulse. Yasuda and Hunter [35] previously reported that deletion of the entire Np24 domain from Pr78 caused a rapid turn over of Pr78 in cells and thus decreased the amount of particles released from cells. It was, therefore suggested that the Np24 domain is important for Pr78 stability. However, deletion of just the KKPKR motif did not alter the intracellular stability of Pr78ΔKR. Instead, the intensity of Pr78ΔKR decreased in the cell lysates in a manner similar to Pr78WT. This was accompanied by a slightly reduced rate of release of virus particles; Pr78WT can first be detected in the medium after 1 h (Fig. 3B), while Pr78ΔKR was first detected in the medium after 2 h (Fig. 3B). Thus, Pr78ΔKR efficiently assembled into procapsids and released processed Gag (p27) in the culture medium with kinetics similar to Pr78WT.
Figure 3 Intracellular procapsid morphology viewed by electron microscopy. COS-1 cells were transfected with pSARM-4 (A and C) or the ΔKKPKR mutant (B and D) proviral DNAs. Wild-type and mutant procapsids observed in close proximity to the nuclear membrane (white arrow) and throughout the cytoplasm near intracellular vesicles (black arrows). Bars approximately 500 nm.
ΔKKPKR intracellular procapsids are indistinguishable from WT procapsids
The metabolic labeling and cell fractionation experiments provided biochemical evidence that the deletion of the KKPKR motif did not affect the ability of Gag to assemble into procapsids and be released from cells. Examining cells expressing Pr78WT and Pr78ΔKR by thin-section EM provided further evidence of normal assembly. Both produce intracellular, spherical procapsids (70–90 nm dia.) that have the characteristic, ring-shaped core typical of immature particles (Fig. 3). In addition, wild-type procapsids were found near the nuclear membrane, which others have shown to be the site of intracellular assembly [19] and near intracellular membranes (Fig. 3A and 3C). Interestingly, Sfakianos and Hunter have previously shown that Pr78WT co-localizes with Rab11+ recycling endosomes [33]. Whether the vesicles shown here are recycling endosomes is not yet known. Of note, we observed fewer assembling, or fully assembled procapsids near the nuclear membrane compared to wild-type (Fig. 3B and 3D) suggesting that at least part of the replication defect may be due to defect in intracellular targeting of newly synthesized Pr78ΔKR proteins.
ΔKKPKR Gag maturation and Envelope packaging
The cell fractionation and pulse-chase experiments combined with the EM analyses showed that the ΔKKPKR deletion mutant was released from cells as virus-like particles. Furthermore, the presence of reverse transcriptase activity and the CA (p27) protein in the culture medium of ΔKKPKR transfected cells suggest that the deletion did not affect PR-mediated processing of Gag, Gag-Pro or Gag-Pro-Pol (Fig. 1 and 2). The other Gag cleavage products (MA, p24/pp18, p12, NC, and p4) were not detected in the [35S] methinione labeling experiments because they do not contain a sufficient number of methionine residues for detection. Furthermore, these experiments could not determine if the viral glycoproteins, SU and TM (gp70 and gp20, respectively) were incorporated into the released particles because an anti-pr78-specific antibody used for the immunoprecipitations. It was therefore possible that the KR box deletion mutation either inhibited Env glycoprotein incorporation or precise Gag processing, thus blocking infectivity. To address this possibility, we looked for the presence of the Env glycoproteins and the other Gag cleavage products in released virions. To this end, COS-1 cells were transfected with pSARM4, pΔKKPKR or an Env-deletion mutant, pMT.ΔE and labeled overnight with [3H] leucine. Viral particles were pelleted through a 25% sucrose cushion, lysed, and immunoprecipitated using an anti-M-PMV antisera that recognizes the Gag cleavage products as well as gp70 and gp20. Figure 4 shows both wild-type and ΔKKPKR mutant particles contain the Env glycoproteins (gp70 and gp20) and the mature Gag cleavage proteins (MA [p10], pp16, p12, and CA [p27]). The NC (p14) and p4 cleavage products were not detected with the antiserum used. As expected particles released from pMT.ΔE transfected cells did not contain the gp70 or gp20 glycoproteins. These results demonstrate that the block to ΔKKPKR replication is not due to abnormal Pr78ΔKR processing or an inability of the mutant particles to incorporate the viral glycoproteins.
Figure 4 Glycoprotein incorporation and Gag processing. COS-1 cells were transfected with nothing (lane 1), pSAMR4 (lane 2), ΔKKPKR (lane 3), or pMT ΔE (lane 4) and then labeled overnight with [3H] leucine. Culture medium was filtered and viral particles were pelleted through a 20% sucrose cushion. The viral proteins present in the pellet were immunoprecipitated using goat anti-M-PMV antibodies. Positions of various viral proteins are indicated.
Genome packaging
Having demonstrated that the deletion of the KR box did not affect the packaging or processing of the gag-, pol-, and env-encoded viral proteins, semi-quantitative RT-PCR assays were utilized to address whether the deletion of the KR box altered packaging of genomic RNA into virions. Equivalent amounts of virus, normalized by p27 content from wild-type and mutant virions were pelleted through a 20% sucrose cushion and resuspended in PBS and the viral RNAs were extracted. Two-fold serial dilutions of viral RNAs were used for RT-PCR reactions using primers to amplify CA-coding sequences. The relative amounts of viral RNA that were packaged were estimated by determining the end-point dilution within which viral cDNAs could be detected by ethidium bromide staining. As shown in this representative experiment, the deletion of the KKPKR motif resulted in a 6–8 fold decrease in genome packaging, relative to wild-type (Fig. 5). Similar results were found using northern blot and dot-blot analyses of vRNAs using a riboprobe specific for M-PMV LTR sequences (data not shown). We concluded from these vRNA packaging assays that the deletion of the KR box significantly reduced the efficiency of vRNA packaging.
Figure 5 RT-PCR analysis of genome packaging in wild-type M-PMV and ΔKKPKR virions. Purified RNA from equivalent amounts of virus was diluted 1:1,000 (lane 3) followed by 2-fold serial dilutions to 1:96,000 (lane 9). First-strand cDNA synthesis was carried out using M-MLV RT and followed by PCR using oligos that amplify M-PMV CA sequences. Relative viral RNA packaging efficiencies were estimated by determining the end-point dilution in which viral PCR products could be detected by Ethidium bromide staining. U, untransfected (lane 1); C, RNA control – no reverse transcriptase added to RT-PCR reaction (lane 2).
Subcellular localization
Electron microscopic examination indicated that fewer numbers of ΔKKPKR procapsids were present near the nuclear membrane and suggested that the deletion of the KR box influences intracellular targeting of Pr78ΔKR to this perinuclear site of assembly. This observation combined with the finding that Pr78ΔKR packages significantly less vRNA into particles lead us to hypothesize that correct intracellular targeting and vRNA packaging are linked. This hypothesis is supported by previous studies with Rous sarcoma virus (RSV) and bovine leukemia virus (BLV) which demonstrated that basic residues in the regions of Gag proteins distant from the RNA binding motif within NC influences both Gag targeting and viral RNA packaging [29,30,32,36].
Studies with RSV have shown that its Gag proteins cycles through the nuclear compartment using a nonclassical nuclear targeting sequence within the MA domain and is exported out of the nucleus via the CRM-1 export pathway. Scheifele et al. have shown that treating the RSV Gag-expressing cells with the CRM-1 inhibitor leptomycin B (LMB) results in a dramatic accumulation of RSV Gag proteins within the nucleus. In addition, it has been shown that RSV MA mutants that are not targeted to the nuclear compartment are insensitive to LMB treatment and are released from cells as virus-like particles, yet are not infectious due to a defect in vRNA packaging. These results suggests that nuclear localization of RSV Gag and genome packaging are linked [29,36]. Likewise, Wang et al. have shown that basic residues within the MA domain of BLV are involved in vRNA packaging [30]. However, BLV Gag was not detected in the nucleus of cells treated with LMV. These results suggest that BLV Gag either does not enter the nucleus or that Gag does enter the nucleus but is exported by a CRM1-independent pathway.
To further explore the intracellular trafficking of Pr78WT and Pr78ΔKR, the steady-state intracellular locations of both were analyzed by confocal microscopy. Figure 6, which are representative z-sections of transfected COS-1 cells, shows that the highest concentration of both Pr78WT and Pr78ΔKR were routinely found throughout the cytoplasm. Interestingly, small amounts of Pr78WT were also observed associated with the nuclear compartment. In contrast, nuclear staining of Pr78ΔKR was only occasionally observed (Figure 6A and 6B, respectively). To examine if M-PMV Gag transiently traffics through the nuclear compartment in a CRM-1-dependent manner similar to RSV, we asked whether M-PMV Gag could be trapped within the nucleus by inhibiting the CRM-1 nuclear export pathway with LMB. As has been previously described [36], RSV Gag-GFP proteins were readily concentrated within the nucleus upon treatment with LMB (Fig. 6E–F). In contrast, LMB did not concentrate either Pr78WT or Pr78ΔKR in the nucleus (Fig. 6A–D).
Figure 6 Subcellular localization of wild-type M-PMV Gag, ΔKKPKR Gag, and RSV Gag-GFP under steady state growth conditions or after treatment with LMB. HeLa cells were transfected with either pSARM-4, ΔKKPKR, or RSV Gag-GFP and left untreated or treated with LMB. The cells were fixed in methanol and the subcellular localizations of Gag were viewed by confocal microscopy using rabbit anti-Pr78 antibodies and Cy2 conjugated secondary antibodies. RSV Gag-GFP was directly visualized by fluorescence of the Gag-GFP fusion protein. Drug treatments: Wild-type M-PMV (untreated, panel 6A), wild-type MPMV (LMB treated, panel 6B) ΔKKPKR (untreated, panel 6C, ΔKKPKR (LMB treated, panel 6D), RSV Gag-GFP (untreated, panel 6E), and RSV Gag-GFP (LMB treated, panel 6F). Colocalization of wild-type M-PMV Gag, ΔKKPKR, and nuclear pores. Transfected HeLa cells were fixed with 4% paraformaldyhyde, and permiablized with 0.2% TX-100. Wild-type Gag (panels G-J), ΔKKPKR Gag (panels k-N), and nuclear pore localization were visualized by confocal microscopy using affinity purified anti-Pr78 and MAb414 antibodies, respectively, and counter stained with Cy2 anti-rabbit and Cy5 anti-mouse antibodies. 0.3 um Z-sections were stacked and orthogonal views through the cell were generated using Flowview imaging analysis software.
Although these immunofluorescence experiments demonstrated that M-PMV does not utilize a CRM-1-dependent nuclear export pathway, it is possible that M-PMV Gag either enters the nucleus, and is exported in a CRM-1-independent manner, or Gag does not enter the nucleus but instead localizes to the nuclear pores. To distinguish between these two possibilities, we utilized confocal microscopy, an affinity-purified rabbit polyclonal anti-Pr78 antibody and a monoclonal antibody (MAb414) that recognizes conserved FG repeats of nuclear pore proteins [37] to examine whether Gag co-localizes with nuclear pores. Following confocal imaging of HeLa cells expressing either Pr78WT or Pr78ΔKR, 0.3 um optical z-sections were stacked and orthogonal views through the nuclei were analyzed. In these cells, the nuclear pores were easily identifiable as punctate staining areas on the nuclear membrane. Moreover, Pr78WT was not randomly dispersed within the nuclear compartment or around the nuclear membrane but rather concentrated in distinct foci in close proximity to nuclear pores (Fig. 6G–J). In contrast Pr78ΔKR did not readily associate with nuclear pores (Fig. 6K–N), but instead localized mainly in the cytoplasm and as discrete foci adjacent to, but not associated with nuclear pores. However, we occasionally, but very infrequently, observed Pr78ΔKR associating with nuclear pores.
Discussion
The results of the studies described here show that a KKPKR sequence located near the carboxy-terminal end of the M-PMV Np24 domain of Gag plays a critical role during virus replication. Initial experiments showed that deletion of this motif did not inhibit the release of virions as demonstrated by the release of virion-associated RT activity into the culture medium of transfected cells. However, this mutant was unable to replicate, similar to what was observed when the entire Np24 domain was deleted [35]. While it is possible that the replication defect was due to the deletion inducing deleterious conformational changes in Pr78, several observations argue that this is unlikely. First, the mutant was capable of assembling into spherical procapsids with morphologies indistinguishable from wild-type. Second, these mutant procapsids, like wild-type, associated with intracellular membranes as has been described as a normal transport pathway for several retroviruses [33,38,39]. Finally, they packaged normal levels of the viral glycoproteins, gp70 and gp20, into the released virions. These results show that the deletion did not significantly alter the confirmation of Pr78 and they demonstrate that the basic residues in Np24 are not involved in targeting Gag to cellular membranes or in packaging the viral glycoproteins.
Further analyses showed that the deletion resulted in two assembly-related defects, vRNA packaging and intracellular targeting. Semi-quantitative PCR, northern blot and dot blot analyses routinely demonstrated that the mutant packaged 6–8 fold less vRNA than did wild-type. In vitro assembly assays have shown that assembly of RSV and HIV-spherical capsids requires nucleic acids [26,40]. In the absence of nucleic acids, Gag proteins assemble into sheets and tubes. Although we have not yet determined whether the spherical, cytoplasmic procapsids assembled from the ΔKKPKR mutant contained RNA (presumably mainly cellular RNAs), we would assume that the RNA content of those intracellular capsids is similar to that found in the released capsids.
Several explanations could account for the failure of this mutant to specifically package vRNAs. First, the Np24 domain may be directly involved in RNA packaging. While there is no evidence that the Np24 domain directly binds RNA or interacts with the NC protein in the virion, the presence of these basic residues may help facilitate NC-mediated vRNA packaging in a manner analogous to that hypothesized for the MA domain of bovine leukemia virus [30].
Another explanation for the defect in genome packaging, although unlikely, is that the deletion disrupted the viral RNA dimerization and/or packaging signals. While it has been proposed that dimer formation is required for RNA packaging, the relationship between dimer formation and packaging is still unclear. Nonetheless, for RSV, MLV, and HIV, the sequence elements involved in dimerization are included in the packaging signals located near the 5'-end of their respective gag genes. In addition, many in vivo studies have shown that mutations that disrupt RNA dimer formation also interfere with RNA packaging. For M-PMV, the RNA dimer initiation sequence (DIS) has not been precisely mapped. However, based on the mapping of the DIS within RNA packaging signals in other retroviruses (reviewed in [41]), the M-PMV DIS is also likely to be an integral part of the RNA packaging signal (Ψ). Because the ΔKKPKR deletion is more than 400 nucleotides down-stream of Ψ packaging signal [42], it is unlikely that the deletion mutation affected the cis-acting elements required for vRNA dimerization or packaging. We are currently determining if the few viral RNAs detected in the ΔKKPKR particles exist as dimmers and whether the mutant vRNA can be packaged in trans with wild-type Gag.
Previous studies on M-PMV suggested that the Np24 domain is required for Pr78 stability. Yasuda and Hunter [35] showed using pulse-chase experiments that deletion of the entire Np24 domain resulted in a Gag protein that when expressed in transfected cells, was more unstable than wild-type. The data presented here suggests that Np24, and perhaps more specifically the KKPKR motif, also plays an important role in intracellular targeting of Pr78. Based on the observations that relatively few ΔKKPKR mutant capsids were found in the perinuclear region of the cytoplasm, which has been shown to be the site of assembly, as well as the findings that wild-type Gag, but not the mutant, localized with the nuclear pores, we hypothesize that the KKPKR motif plays a role in targeting Gag to the site of genome packaging, which may be either be at the nuclear pores, within the nucleus, as suggested by Scheifele et al. [36], or in the cytoplasm juxtaposed to the outer nuclear membrane.
Interestingly, the KKPKR motif resembles a classical nuclear localization signal [43] and we have previously found that over expression of the nuclear pore associated, Ubc9 protein lead to a dramatic redistribution of Pr78 to the nuclear compartment [44]. Experiments are in progress to determine whether the KKPKR motif can function as a nuclear targeting signal when fused to a heterologous protein. If the KKPKR does function as an NLS to cycle Pr78 through the nucleus during the virus life cycle, as has been suggested for RSV, its export to the cytoplasm does not utilize the CRM-1 pathway (Fig. 6B). An alternative hypothesis, which is consistent with the data presented here, is that the KKPKR sequence targets Pr78 to the nuclear pore, where it first recognizes the vRNA during Tap-mediated RNA export [45]. This Gag-vRNA complex would then serve as the nucleation event for spherical capsid assembly just outside of the nuclear pore where the betaretroviruses are known to assemble.
A motif within the M-PMV MA domain called the cytoplasmic targeting/retention signal (CTRS), which is located approximately 100 residues upstream of the KKPKR motif, has also been implicated in directing Pr78 to the intracellular site of assembly. Mutant Pr78 proteins (R55W) that contain an arginine to tryptophan substitution at position 55 in MA do not accumulate at the usual cytoplasmic sites of assembly. Instead R55W-Pr78 proteins are targeted the plasma membrane where they assemble concomitantly with budding, as with the C-type retroviruses [20]. This arginine is contained within an 18 amino acid sequence (residues 43–60) that is conserved between M-PMV and MMTV. When these 18 residues were inserted in the MA domain of C-type MLV Gag, MLV capsid assembly occurred in the cytoplasm [46]. Whether or not these altered MLV capsids assembled in the perinuclear/pericentriolar region of the cytoplasm was not shown. It has, therefore, been suggested that these residues either target Pr78 molecules to the cytoplasmic assembly site or they retain Pr78 at this site until the procapsids are fully assembled.
We hypothesize that the KKPKR motif identified in this study, which may be included in a larger motif that has yet to fully defined, functions either prior to or separate from the CTRS function. We speculate that the KKPKR motif is involved in targeting Pr78 to the nuclear pore to facilitate RNA packaging. Because only two copies of vRNA are packaged into virions, only a few Gag proteins need to be targeted there, which is consistent with the findings present here that the majority of Gag proteins don't associate with nuclear pores. Once Gag has associated with the vRNA, the Gag-vRNA complex would then be transported to the assembly site perhaps via the CTRS signal to initiate spherical capsid assembly.
Materials and methods
DNAs
Plasmid pSARM4 is an infectious molecular clone of wild-type M-PMV. Plasmid pMT.ΔE is an env deletion mutant of pSARM4 [47]. Deletion of the KKPKR motif was accomplished using the Altered Sites II Mutagenesis System (Promega) as per manufacture's protocol. Briefly, the 1,307 bp, SphI-PvuII fragment (nt 171-1478) of pSARM4 was subcloned into the SmaI-SphI sites of pALTER. Mutagenesis was carried out using the mutagenic oligonucleotide (5'-GTTTGTGCTCTTAACAGAACT GGGAAAGTACTTGATAAACCTTTATCTTGTAGAGAGG), to precisely delete amino acids 153 through 157 (KKPKR) in M-PMV Gag. The mutation was subcloned back into pSARM4 using SacI and PacI sites. After mutagenesis, plasmid DNAs were sequenced to ensure that unwanted mutations were not inadvertently created. Plasmid pETM100A is a prokaryote expression vector used to express a (His)6-tagged M-PMV Gag protein in E. coli (32). Plasmid pRS.V8-EGFP was used to express a RSV Gag-EGFP fusion protein in mammalian cells (John Wills, Pennsylvania State University College of Medicine) [29].
Cell lines and transfection
COS-1 and HOS cells were grown at 37°C with 5% CO2 in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum. HeLa cells were grown in RPMI 1640 medium supplemented with 10% fetal bovine serum and 5% tryptose phosphate broth. DNA transfections were carried out using Fugene 6 (Roche Diagnostics, Indianapolis, IN) following the manufacture's protocol.
Antibodies
Goat anti-M-PMV antibodies were obtained from Eric Hunter (Emory University). Mouse monoclonal antibodies that recognize the conserved FG repeats found in nuclear pore complex proteins (MAb414) were purchased from Covance Research Products (Berkely, CA). HRP-conjugated goat anti-rabbit IgG and HRP-conjugated goat anti-mouse IgG were purchased from Amersham Pharmacia Biotech (Little Chalfont Buckinghamshire, England). Cy2 conjugated donkey anti-rabbit and Cy5 conjugated donkey anti-mouse were purchased from Jackson Immunoresearch Laboratories (West Grove, PA). Rabbit polyclonal anti-Pr78 (47) was affinity purified as follows. M-PMV Gag proteins containing a carboxy-terminal (His)6 tag were expressed in E. coli BL21 (DE3) cells from the expression plamsid pET.M100A for 4 hours in the presence of 0.1 mM IPTG. Cells were harvested by centrifugation at 4,000 × g and lysed at room temperature in a denaturing lysis buffer containing 8 M urea in TNI pH 8.0 buffer (50 mM Tris Cl pH 8.0, 150 mM NaCl and 0.5 mM imidizole). Cellular debris was removed by centrifugation at 10,000 × g and the supernatant was passed over an Ni-NTA agarose nickel column (Qiagen Sciences, Maryland, USA) and extensively washed in 8 M urea in TNI pH 6.5 buffer. The denatured Gag proteins on the column were refolded using a slow (8 h), linear reduction of urea from 8 M to 0 M (in TNI buffer, pH 8.0) in a manner similar to that described by Klikova et al. [48] who have shown that M-PMV Gag proteins denatured in 8 M urea could be refolded into a confirmations that are competent to assemble into immature capsids in vitro by slowly removing the urea over an 8 hour period at 4°C. After refolding, Gag proteins were eluted form the Ni2+ column using TNI pH 8.0 buffer containing 0.1 mM EDTA. After exchanging the TNI-EDTA buffer to pH7.2 Coupling Buffer (Pierce) by dialysis, the Gag proteins were covalently coupled to activated agarose beads using the Amino-Link Plus kit (Pierce, Rockford, Il) per manufacturer's protocol. Anti-Pr78 antibodies in the immunized rabbit serum were affinity purified using the immobilized Gag column by standard protocols [49].
Intracellular Gag Fractionation and Immunoblots
Intracellular Gag fractionation experiments were used to assay intracellular procapsid assembly. 60 mm diameter culture dishes containing either untransfected or transfected COS-1 cells were lysed in 1 ml TX-100 lysis buffer (0.25 M sucrose, 1.0 mM EDTA, 10 mM Tris-HCl [pH7.5], 0.14 M NaCl, 0.5% Triton X-100, 0.25% DOC) containing PMSF, leupeptin, pepstatin, and aprotinin for 30 minutes on ice. Cell debris and nuclei were removed by centrifugation for 2 min in a microfuge. The pellets were discarded and the capsid-containing supernatants were fractionated by centrifugation through a 20% sucrose cushion at 350,00 × g in a MLA-130 rotor for 30 minutes at 4°C. The pelleted procapsids were lysed in 2× protein loading buffer (10% glycerol, 2.3% SDS, 63 mM Tris-HCL [pH6.8], 5% β-mercaptoethanol, and 0.01% bromophenol blue). Lysates were boiled for 3 minutes, separated on a 12% SDS PAGE gel, transferred to nitrocellulose. Viral proteins were analyzed by western blot using polyclonal anti-Pr78, HRP-conjugated goat anti-rabbit IgG, and Western Lightning Chemiluminescence Reagent Plus (Perkin-Elmer) as per manufacturers suggestions.
Radiolabeling and immunoprecipitation
For this, subconfluent 60 mm diameter tissue culture dishes containing either untransfected and transfected COS-1 cells were washed twice in DMEM without methionine or cysteine (DMEM Met- Cys-) and starved in 4 ml of DMEM Met- Cys- for 15 min at 37°C in 5% CO2. The medium was replaced with 800 μl of DMEM Met- Cys- containing 250 μCi [35S] methionine-cysteine (>1000 Ci/mmol; NEN, Boston, MA). The cells were incubated for 30 minutes at 37°C, 5% CO2. Cells were either immediately lysed (pulse) or washed with complete medium and then incubated for 1, 2, 3, 4, or 8 hours in complete medium (chase) prior to lysis.
Virion and cell lysis were lysed as follows. Culture medium from each plate was collected and clarified by centrifugation in a microfuge for 2 min at 13,000 rpms. Virions were lysed by adjusting the culture medium to 1× lysis buffer B by adding 1/5 volume of 5× lysis buffer B (0.5% sodium dodecyl sulfate [SDS], 5% Triton X-100, 5% deoxycholate [DOC], 0.75 M NaCl, 0.25 M Tris-HCl [pH 6.8]). Cell monolayers were lysed in 1 ml lysis buffer A (Triton X-100, 1% DOC, 0.15 M NaCl, 0.05 M Tris-HCl [pH 6.8]) containing PMSF, leupeptin, pepstatin, and aprotinin for 30 minutes on ice. Cell debris and nuclei were removed by centrifugation for 2 min in a microfuge. The pellets were discarded and the capsid-containing supernatants were adjusted to 1× lysis buffer B by adding 0.1% SDS. All lysates were precleared by incubation with inactivated, formalin-fixed Staphylococcus aureus cells. Viral proteins were immunoprecipitated from all samples using Rbt anti-Pr78 antibodies and Staphylococcus aureus cells as previously described [50]. Immunoprecipitates were resuspended in protein loading buffer (10% glycerol, 2.3% SDS, 63 mM Tris-HCL [pH6.8], 5% β-mercaptoethanol, and 0.01% bromophenol blue), boiled for 3 minutes, separated by SDS PAGE gel (12% acrylamide), and visualized by fluorography or analyzed by phosphor-imaging using The Discovery Series Quantity One (BioRad, Hercules, CA).
Steady state radiolabeling with [3H] leucine was done to assess glycoprotein incorporation and Pr78 processing. Transfected COS-1 cells were starved in leucine-free DMEM for 90 minutes, and labeled overnight with [3H] leucine (500 μCi/ml, 173.0 Ci/mmol, NEN, Boston, MA). The culture medium was filtered with a 0.45 μm syringe filter, and [3H] leucine-labeled viruses were pelleted through a 25% sucrose cushion by centrifugation at 350,000 × g in a TLA 100.3 rotor for 30 minutes at 4°C. After the viral pellet was solublized in 1× lysis buffer B, the viral proteins were using goat anti-M-PMV antibodies and analyzed as described above.
Virus replication assay
The infectivity of wild-type and mutant viruses were determined by measuring the increase of reverse transcriptase (RT) activity in the culture supernatants of inoculated HOS cell cultures at various times postinfection. Culture medium was harvested from COS-1 cells that had been transfected 48 h previously with either wild-type or mutant DNA. Loose cells and cellular debris were pelleted by centrifugation for 2 min in a microfuge and the level of RT activity in the clarified culture medium was measured. Hos cells were infected with equivalent amount of RT-containing in the presence of 4.0 μg/ml of polybrene. Culture fluids were harvested at various days postinfection and assayed for RT activity. RT assays were carried out by pelleting virus from medium by centrifugation at 350,000 × g in a TLA 100.3 rotor for 30 minutes at 4°C. The viral pellet was lysed in 12 μl of Virion Lysis Buffer (50 mM Tris-HCl [pH 7.8], 100 mM KCl, 0.05% TX-100, 2 mM diothiothreitol [dTT]) on ice for 15 min. 7.5 μl of virion lysates were added to 30 μl of RT Reaction Buffer (50 mM Tris [pH 8.0], 100 mM KCl, 2 mM dTT, 7.5 mM MgCl2) with 8 μCi [32P] α-TTP (>1000 mCi/mmol, NEN) and 1.25 μg poly (A) – oligo (dT)15 (Roche). The RT reaction was incubated at 37°C for 2 hours. 10 μl of the RT reaction was spotted onto a piece of Whatman DE81 paper and allowed to dry. The filter paper was washed twice for 15 min in 2× SSC buffer (0.3 M NaCl, 0.03 M Na Citrate), twice briefly in 95% EtOH, and once in distilled water. The filters were dried and [32P] incorporation was measured by scintillation counting.
RNA extraction and RT-PCR
Medium from transfected COS-1 cells were clarified and virus was pelleted through a 20% sucrose cushion at 207,570 × g in an SW41 rotor for 2 hours at 4°C and resuspended in 30 μl of PBS. The amount of viral particles was normalized by quantitation of p27 detected by immunobloting. RNA was extracted from equivalent amounts of virus using QIAamp Viral RNA Mini Kit (Qiagen Sciences, Maryland, USA) as per manufacturers suggestions. Purified RNA was treated with 1 U of Rnase-free Dnase I (New England Biolabs, Inc., Berverly, MA) for 30 min at 37°C, followed by inactivation at 70°C for 30 min. Purified RNA from equivalent amounts of virus was diluted 1:1,000 followed by 2-fold serial dilutions. 5 μl of diluted RNA was used for first-strand cDNA synthesis as per manufacturers suggestions for M-MLV RT (Invitrogen) using 500 ng of oligo (dT)12–18 (Ambion, Austin, TX). First-strand cDNA (5 μl) were amplified by PCR using the following oligos 5'-CCGCTCGAGCGGGCCGCCATGCCGGTGGCTGAAACCGTTG and 5'-GCTCTAGAGCGGCGGCCATGGCCAGG to amplify M-PMV CA sequences. The relative amount of viral RNA packaged into viral particles was estimated by end-point dilution.
Electron microscopy
Transmission Electron Microscopy (TEM) was utilized to view assembled intracellular procapsids. Transfected COS-1 cells were fixed for 1 hour with two changes of 3% glutaraldehyde in 0.1 M phosphate buffer (5 mM NaH2PO4 and 5 mM phosphoric acid). The cells were rinsed for 30 min with 0.1 M phosphate buffer, followed by osmication (2% OsO4 in phosphate buffer) for 1 hour. The cultures were washed and dehydrated with a graded series of ethanol (25% 50%, 75%, 95% 100), followed by a graded series of ethanol/Epon 812 (Shell) mixtures (3:1, 1:1, 1:3). The cells were infiltrated with pure Epon 812 and polymerized at 60°C for 48 hours. Thin sections were made with a LBK Ultrotome III, mounted on copper grids, and stained with 2% uranyl acetate and lead citrate. Thin sections were examined and photographed with a HitachiH-7500 transmission electron microscope operated at 60 Kv.
Subcellular localization
The intracellular localizations of Pr78Gag and Pr78ΔKR proteins were determined by confocal microscopy. Transfected HeLa cells were grown on sterile coverslips in 35 mm culture dishes. At 48 hours post transfection, cells were washed with PBS (137 mM NaCl2, 2.7 mM KCl2, and 8 mM Na2HPO4, 2 mM KH2PO4), fixed in either 4% paraformaldyhyde in PBS for 20 minutes at room-temperature and then subsequently permiabilized with 0.2% TX-100 in PBS for 5 minutes at room-temperature, or fixed in 100% methanol at -20°C for 10 minutes. The coverslips were washed with PBS, and blocked with Blocking Buffer 1 (PBS, 0.2% tween 20, 0.4% fish skin gelatin [Sigma]) for 5 minutes, and then blocked with Blocking Buffer 2 (PBS, 0.2% tween 20, 2.5% goat serum [Sigma]) for 5 min. Affinity purified anti-Pr78 and MAb414 (in PBS, 0.2% tween 20, 2.5% goat serum) were incubated on the coverslips for 45 min at 37°C. The primary antibodies were removed and the coverslips were blocked as they had been previously. Cy2 and Cy5 conjugated secondary antibodies (in PBS, 0.2% tween 20, 2.5% goat serum) were incubated on the cover-slips for 30 min at 37°C and washed in PBS with 0.2% tween 20. The coverslips were mounted on slides using GEL-mount and analyzed by confocal microscopy (Olympus FV500 w/upright BX Olympus florescence microscope). 0.3 μm Z-sections were stacked and orthogonal views through the cell were generated using Flowview imaging analysis software (Olympus).
Acknowledgements
We thank Tareq Jaber, Gentry Rundle, Gopinath Seetharaman, John West, and Charles Wood for thoughtful reviews of this manuscript. We also thank Terri Fangman at the University of Nebraska Microscopy Core Facility for her help with confocal imaging. This project described was supported by NIH Grant Number P20 RR15635 from the COBRE Program of the National Center for Research Resources.
==== Refs
Baillie GJ van de Lagemaat LN Baust C Mager DL Multiple groups of endogenous betaretroviruses in mice, rats, and other mammals J Virol 2004 78 5784 5798 15140976 10.1128/JVI.78.11.5784-5798.2004
Hook LM Agafonova Y Ross SR Turner SJ Golovkina TV Genetics of mouse mammary tumor virus-induced mammary tumors: linkage of tumor induction to the gag gene J Virol 2000 74 8876 8883 10982330 10.1128/JVI.74.19.8876-8883.2000
Xu L Sakalian M Shen Z Loss G Neuberger J Mason A Cloning the human betaretrovirus proviral genome from patients with primary biliary cirrhosis Hepatology 2004 39 151 156 14752833 10.1002/hep.20024
Power MD Marx PA Bryant ML Gardner MB Barr PJ Luciw PA Nucleotide sequence of SRV-1, a type D simian acquired immune deficiency syndrome retrovirus Science 1986 231 1567 1572 3006247
Sonigo P Barker C Hunter E Wain-Hobson S Nucleotide sequence of Mason-Pfizer monkey virus: an immunosuppressive D-type retrovirus Cell 1986 45 375 385 2421920 10.1016/0092-8674(86)90323-5
Thayer RM Power MD Bryant ML Gardner MB Barr PJ Luciw PA Sequence relationships of type D retroviruses which cause simian acquired immunodeficiency syndrome Virology 1987 157 317 329 2435057 10.1016/0042-6822(87)90274-1
Colcher D Heberling RL Kalter SS Schlom J Squirrel monkey retrovirus: an endogenous virus of a new world primate J Virol 1977 23 294 301 69724
Carlson J Lyon M Bishop J Vaiman A Cribiu E Mornex JF Brown S Knudson D DeMartini J Leroux C Chromosomal distribution of endogenous Jaagsiekte sheep retrovirus proviral sequences in the sheep genome J Virol 2003 77 9662 9668 12915578 10.1128/JVI.77.17.9662-9668.2003
Cousens C Minguijon E Dalziel RG Ortin A Garcia M Park J Gonzalez L Sharp JM de las Heras M Complete sequence of enzootic nasal tumor virus, a retrovirus associated with transmissible intranasal tumors of sheep J Virol 1999 73 3986 3993 10196294
Kajikawa O Dahlberg JE Rosadio RH De Martini JC Detection and quantitation of a type D retrovirus gag protein in ovine pulmonary carcinoma (sheep pulmonary adenomatosis) by means of a competition radioimmunoassay Vet Microbiol 1990 25 17 28 2247933 10.1016/0378-1135(90)90089-E
Baillie GJ Wilkins RJ Endogenous type D retrovirus in a marsupial, the common brushtail possum (Trichosurus vulpecula) J Virol 2001 75 2499 2507 11160757 10.1128/JVI.75.5.2499-2507.2001
Mager DL Freeman JD Novel mouse type D endogenous proviruses and ETn elements share long terminal repeat and internal sequences J Virol 2000 74 7221 7229 10906176 10.1128/JVI.74.16.7221-7229.2000
Ristevski S Purcell DF Marshall J Campagna D Nouri S Fenton SP McPhee DA Kannourakis G Novel endogenous type D retroviral particles expressed at high levels in a SCID mouse thymic lymphoma J Virol 1999 73 4662 4669 10233925
Chopra HC Bogden AE Zelljadt I Jensen EM Virus particles in a transplantable rat mammary tumor of spontaneous origin Eur J Cancer 1970 6 287 290 5485264
Fine DL Landon JC Pienta RJ Kubicek MT Valerio MG Loeb WF Chopra HC Responses of infant rhesus monkeys to inoculation with Mason-Pfizer monkey virus materials J Natl Cancer Inst 1975 54 651 658 804565
Marx PA Maul DH Osborn KG Lerche NW Moody P Lowenstine LJ Henrickson RV Arthur LO Gilden RV Gravell M Simian AIDS: isolation of a type D retrovirus and transmission of the disease Science 1984 223 1083 1086 6695196
Stromberg K Benveniste RE Arthur LO Rabin H Giddens WEJ Ochs HD Morton WR Tsai CC Characterization of exogenous type D retrovirus from a fibroma of a macaque with simian AIDS and fibromatosis Science 1984 224 289 282 6200929
Freed EO HIV-1 gag proteins: diverse functions in the virus life cycle Virology 1998 251 1 15 9813197 10.1006/viro.1998.9398
Sfakianos JN LaCasse RA Hunter E The M-PMV cytoplasmic targeting-retention signal directs nascent Gag polypeptides to a pericentriolar region of the cell Traffic 2003 4 660 670 12956869 10.1034/j.1600-0854.2003.00125.x
Rhee SS Hunter E A single amino acid substitution within the matrix protein of a type D retrovirus converts its morphogenesis to that of a type C retrovirus Cell 1990 63 77 86 2170021 10.1016/0092-8674(90)90289-Q
Rhee SS Hunter E Amino acid substitutions within the matrix protein of type D retroviruses affect assembly, transport and membrane association of a capsid Embo J 1991 10 535 546 1705884
Rhee SS Hunter E Structural role of the matrix protein of type D retroviruses in gag polyprotein stability and capsid assembly J Virol 1990 64 4383 4389 2200887
Strambio-de-Castillia C Hunter E Mutational analysis of the major homology region of Mason-Pfizer monkey virus by use of saturation mutagenesis J Virol 1992 66 7021 7032 1279197
Sakalian M Hunter E Separate assembly and transport domains within the Gag precursor of Mason-Pfizer monkey virus J Virol 1999 73 8073 8082 10482556
Zabransky A Sakalian M Pichova I Localization of self-interacting domains within betaretrovirus Gag polyproteins Virology 2005 332 659 666 15680431 10.1016/j.virol.2004.12.007
Campbell S Rein A In vitro assembly properties of human immunodeficiency virus type 1 Gag protein lacking the p6 domain J Virol 1999 73 2270 2279 9971810
Campbell S Vogt VM Self-assembly in vitro of purified CA-NC proteins from Rous sarcoma virus and human immunodeficiency virus type 1 J Virol 1995 69 6487 6497 7666550
Dupont S Sharova N DeHoratius C Virbasius CM Zhu X Bukrinskaya AG Stevenson M Green MR A novel nuclear export activity in HIV-1 matrix protein required for viral replication Nature 1999 402 681 685 10604476 10.1038/45272
Callahan EM Wills JW Link between genome packaging and rate of budding for Rous sarcoma virus J Virol 2003 77 9388 9398 12915554 10.1128/JVI.77.17.9388-9398.2003
Wang H Norris KM Mansky LM Involvement of the matrix and nucleocapsid domains of the bovine leukemia virus Gag polyprotein precursor in viral RNA packaging J Virol 2003 77 9431 9438 12915558 10.1128/JVI.77.17.9431-9438.2003
Parent LJ Cairns TM Albert JA Wilson CB Wills JW Craven RC RNA dimerization defect in a Rous sarcoma virus matrix mutant J Virol 2000 74 164 172 10590103
Garbitt RA Albert JA Kessler MD Parent LJ trans-acting inhibition of genomic RNA dimerization by Rous sarcoma virus matrix mutants J Virol 2001 75 260 268 11119596 10.1128/JVI.75.1.260-268.2001
Sfakianos JN Hunter E M-PMV capsid transport is mediated by Env/Gag interactions at the pericentriolar recycling endosome Traffic 2003 4 671 680 12956870 10.1034/j.1600-0854.2003.00126.x
Gottwein E Bodem J Muller B Schmechel A Zentgraf H Krausslich HG The Mason-Pfizer monkey virus PPPY and PSAP motifs both contribute to virus release J Virol 2003 77 9474 9485 12915562 10.1128/JVI.77.17.9474-9485.2003
Yasuda J Hunter E A proline-rich motif (PPPY) in the Gag polyprotein of Mason-Pfizer monkey virus plays a maturation-independent role in virion release J Virol 1998 72 4095 4103 9557699
Scheifele LZ Garbitt RA Rhoads JD Parent LJ Nuclear entry and CRM1-dependent nuclear export of the Rous sarcoma virus Gag polyprotein Proc Natl Acad Sci U S A 2002 99 3944 3949 11891341 10.1073/pnas.062652199
Davis LI Blobel G Identification and characterization of a nuclear pore complex protein Cell 1986 45 699 709 3518946 10.1016/0092-8674(86)90784-1
Kramer B Pelchen-Matthews A Deneka M Garcia E Piguet V Marsh M HIV interaction with endosomes in macrophages and dendritic cells Blood Cells Mol Dis 2005 35 136 142 16087369 10.1016/j.bcmd.2005.06.006
Nydegger S Foti M Derdowski A Spearman P Thali M HIV-1 egress is gated through late endosomal membranes Traffic 2003 4 902 910 14617353 10.1046/j.1600-0854.2003.00145.x
Campbell S Vogt VM In vitro assembly of virus-like particles with Rous sarcoma virus Gag deletion mutants: identification of the p10 domain as a morphological determinant in the formation of spherical particles J Virol 1997 71 4425 4435 9151833
Paillart JC Shehu-Xhilaga M Marquet R Mak J Dimerization of retroviral RNA genomes: an inseparable pair Nat Rev Microbiol 2004 2 461 472 15152202 10.1038/nrmicro903
Schmidt RD Mustafa F Lew KA Browning MT Rizvi TA Sequences within both the 5' untranslated region and the gag gene are important for efficient encapsidation of Mason-Pfizer monkey virus RNA Virology 2003 309 166 178 12726736 10.1016/S0042-6822(02)00101-0
Silver PA How proteins enter the nucleus Cell 1991 64 489 497 1991319 10.1016/0092-8674(91)90233-O
Weldon RAJ Sarkar P Brown SM Weldon SK Mason-Pfizer monkey virus Gag proteins interact with the human sumo conjugating enzyme, hUbc9 Virology 2003 314 62 73 14517060 10.1016/S0042-6822(03)00348-9
Braun IC Rohrbach E Schmitt C Izaurralde E TAP binds to the constitutive transport element (CTE) through a novel RNA-binding motif that is sufficient to promote CTE-dependent RNA export from the nucleus Embo J 1999 18 1953 1965 10202158 10.1093/emboj/18.7.1953
Choi G Park S Choi B Hong S Lee J Hunter E Rhee SS Identification of a cytoplasmic targeting/retention signal in a retroviral Gag polyprotein J Virol 1999 73 5431 5437 10364290
Song C Hunter E Variable sensitivity to substitutions in the N-terminal heptad repeat of Mason-Pfizer monkey virus transmembrane protein J Virol 2003 77 7779 7785 12829817 10.1128/JVI.77.14.7779-7785.2003
Klikova M Rhee SS Hunter E Ruml T Efficient in vivo and in vitro assembly of retroviral capsids from Gag precursor proteins expressed in bacteria J Virol 1995 69 1093 1098 7815488
Harlow ELD Antibodies; A laboratory manual 1988 Cold Spring Harbor , Cold Spring Harbor Laboratory Press 726
Sakalian M Parker SD Weldon RAJ Hunter E Synthesis and assembly of retrovirus Gag precursors into immature capsids in vitro J Virol 1996 70 3706 3715 8648705
|
16274484
|
PMC1308863
|
CC BY
|
2021-01-04 16:36:39
|
no
|
Retrovirology. 2005 Nov 7; 2:68
|
utf-8
|
Retrovirology
| 2,005 |
10.1186/1742-4690-2-68
|
oa_comm
|
==== Front
RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-691628007610.1186/1742-4690-2-69ReviewTat gets the "green" light on transcription initiation Brady John [email protected] Fatah [email protected] National Cancer Institute, Laboratory of Cellular Oncology, Bethesda, MD 20892, USA2 The George Washington University School of Medicine, Department of Biochemistry and Molecular Biology, Washington, DC 20037, USA2005 9 11 2005 2 69 69 28 9 2005 9 11 2005 Copyright © 2005 Brady and Kashanchi; licensee BioMed Central Ltd.2005Brady and Kashanchi; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Human immunodeficiency virus type 1 (HIV-1) Tat transactivation is an essential step in the viral life cycle. Over the past several years, it has become widely accepted that Tat exerts its transcriptional effect by binding the transactivation-responsive region (TAR) and enhancing transcriptional elongation. Consistent with this hypothesis, it has been shown that Tat promotes the binding of P-TEFb, a transcription elongation factor composed of cyclin T1 and cdk9, and the interaction of Tat with P-TEFb and TAR leads to hyperphosphorylation of the C-terminal domain (CTD) of RNA Pol II and increased processivity of RNA Pol II. A recent report, however, has generated renewed interest that Tat may also play a critical role in transcription complex (TC) assembly at the preinitiation step. Using in vivo chromatin immunoprecipitation assays, the authors reported that the HIV TC contains TBP but not TBP-associated factors. The stimulatory effect involved the direct interaction of Tat and P-TEFb and was evident at the earliest step of TC assembly, the TBP-TATA box interaction. In this article, we will review this data in context of earlier data which also support Tat's involvement in transcriptional complex assembly. Specifically, we will discuss experiments which demonstrated that Tat interacted with TBP and increased transcription initiation complex stability in cell free assays. We will also discuss studies which demonstrated that over expression of TBP alone was sufficient to obtain Tat activated transcription in vitro and in vivo. Finally, studies using self-cleaving ribozymes which suggested that Tat transactivation was not compatible with pausing of the RNA Pol II at the TAR site will be discussed.
==== Body
Tat transactivation: A historical perspective, initiation vs elongation
Transcription of the HIV-1 provirus is characterized by an early, Tat-independent and a late, Tat-dependent phase. Transcription from the HIV-1 LTR is increased several hundred-fold in the presence of Tat and the ability of Tat to activate transcription is essential for virus replication. Tat is an unusual transcription factor because it interacts with a cis acting RNA enhancer element, TAR, present at the 5' end of all viral transcripts (nt +1 to +59) [1-4]. In fact, TAR was the first demonstration of a RNA enhancer element. Unlike other eukaryotic enhancers, however, the TAR element was only functional when it was placed 3' to the HIV promoter and in the correct orientation and position [5]. The location of the TAR in transcribed regions was surprising, and to many, inconsistent with a role for TAR in transcription initiation. In fact, the uniqueness of the RNA enhancer element drove many investigators to search for unique pathways in HIV Tat transactivation. When Kao et al. [6] reported that in the absence of Tat the majority of RNA polymerases initiating transcription stall near the promoter, and later Laspia et al. [7] reported a small effect of Tat on transcription initiation but a large effect on transcription elongation, the initiation model quickly lost support. The observation that Tat binds specifically to the TAR RNA [8] and could function as an RNA binding protein [9] gave further support for the elongation model, and it became quite well accepted that through interaction with TAR, Tat promotes the assembly of an active transcription elongation complex. The more recent finding that Tat promotes the binding of P-TEFb, a transcription elongation factor composed of cyclin T1 and cdk9 [10] and, more recently, Brd4 in the active nuclear complex [11] seemed consistent with the elongation model. In fact, it has been shown that the interaction of Tat with P-TEFb and TAR leads to hyperphosphorylation of the C-terminal domain (CTD) of RNA Pol II and increased processivity of RNA Pol II [12-22]. Moreover, Tat induces P-TEFb dependent phosphorylation of Tat-SF1 and SPT5 [23]. While TAR plays a critical role in Tat transactivation, it is also clear that optimal Tat transactivation of HIV-1 gene expression requires upstream transcription co-factors. Along these lines, it has been reported that Tat physically interacts with the pre-initiation complex including transcription factors such as Sp1 [24], TATA binding protein (TBP) [25-27], cylinE/cdk2 [28], TFIIH [21,22], Tip60 [29], RNA Pol II [30,31], as well as coactivators such as CBP/p300 [32] and p/CAF [33,34]. Several excellent reviews of the role of Tat in transactivation have been published [1,35-40].
A role for Tat in transcription preinitiation complex assembly
A recent report from M. Green's lab has, however, generated renewed interest that Tat's primary effect may in fact be at the transcription complex (TC) assembly stage at the pre-initiation step upstream of the +1 area, thereby promoting both transcription initiation and elongation of HIV-1 promoter [41]. The authors reported that Tat stimulates TC assembly through a TAF-less TBP complex, thereby promoting initiation and elongation [41]. The stimulatory effect was evident at the earliest step of TC assembly, the TBP-TATA box interaction. Furthermore, much like the scenario in yeast, transcription of protein-coding genes may involve alternative TCs that differ by the presence or absence of certain TAFs. To analyze transcription stimulation by Tat and other activators, such as VP16 and E1A, they performed ChIP experiments in transiently transfected mammalian cells. Following addition of Tat, there was a large increase in association of TBP, TFIIB, mediator (enabling transcriptional activators to regulate transcription by RNA polymerase II), and RNA polymerase II with the promoter. The increased binding of these basal transcription factors paralleled the increase in transcription. Interestingly, although TBP and the other GTFs were efficiently recruited to the promoter in the presence of Tat, there was no significant recruitment of the two TAFs analyzed, TAF1 (TAFII250) or TAF5 (TAFII100). Consistent with their transfection data, they observed the presence of TBP, TFIIB, mediator, Sp1, P-TEFb and RNA polymerase II with the integrated proviral promoter in chronically HIV-1-infected cell lines, 8E5/LAV and U1.
In parallel control ChIP experiments analyzing Gal4-VP16 and Gal4-E1a, the investigators demonstrated that these activators supported assembly of a transcription complex that contained all of the GTFs, including TAFII250 and TAFII100. By contrast, Gal4-Tat directed assembly of a TC in which the TAFs were present at levels significantly below that of TBP and other GTFs. Remarkably, when assaying for effect of cyclin T1 and CDK9, they observed that P-TEFb was responsible for recruitment of this unique TBP complex.
Finally, they concluded that RNA polymerase II was not detected either near or far downstream of the transcription start site in the absence of Tat and thus provided no evidence for a paused (or stalled) RNA polymerase II. Consistent with their ChIP data, nuclear run-off experiments showed that Tat increased the density of RNA polymerase II 9- to 15-fold within the first 25 nucleotides downstream of the transcription start site, indicating that Tat stimulates initiation.
It is interesting to note that while the authors do not see a dependency on TAFII250 for Tat transactivation on the HIV LTR, the interaction of Tat and TAFII250 is important for Tat-mediated transcription repression. Tat represses transcription of both the MHC class I genes and the beta2-microglobulin gene. Repression results from the interaction of Tat with the TAF1 component of the general transcription factor, TFIID and depends exclusively on the C-terminal domain of Tat, beginning at amino acid 73, with a C-terminal limit between amino acids 80 and 83. Tat repressor function also depends on the presence of a lysine at position 41, located within the core of the protein. Tat repressor activity is independent of two N-terminal domains essential for transactivation: the acidic segment and the cysteine-rich region. The C-terminal domain of Tat binds to a site on TAF1 that overlaps the acetyl transferase (AT) domain, inhibiting TAF1 acetyl transferase (AT) activity. Furthermore, promoters repressed by Tat, including the MHC class I promoter, are dependent on TAF1 whereas those that are not repressed by Tat, such as SV40 and MuLV promoters, are independent of functional TAF1 [42-45].
Further evidence for the role of Tat in preinitiation complex assembly
While these studies have renewed interest in the role of Tat in promoting transcription initiation, the idea is certainly not new. For instance, Kashanchi et al. reported in 1994 that the transcriptional activity of HeLa extracts were depleted after chromatography on a Tat affinity column, through specific retention of TBP and some TAFs. The core domain of Tat, amino acids 36–50, was required for the interaction of Tat with TBP and a mutation at Lys 41, which abolishes transactivation, abolished interaction with TBP [46,47]. In fact, based on these results and earlier studies that Tat increased transcription initiation complex stability in cell free assays [48], the authors speculated in this paper that Tat may increase the association or dissociation of TFIID, or recruit a particular species of functionally different TFIID to the HIV template. In contrast to the studies of Raha et al. [41], by western blot analysis Kashanchi et al. [46] detected TAFII250 in the Tat-induced transcription complex. The relative abundance of TBP and TAFII250 was not quantitatively evaluated, however, so it is possible that less TAFII250 was associated with the Tat-TBP complex.
Other studies have also pointed toward the functional interaction of Tat and TBP. The activity of Tat, either wild-type or fused to the DNA binding domain of GAL4 (GBTat), was tested using reporter constructs containing GAL4 binding sites upstream of a minimal promoter corresponding to the HIV-1 TATA box, with or without the TAR element. Overexpression of TBP led to a dramatic increase in the activity of the GBTat protein. Analysis of several Tat mutants indicated that both the cysteine-rich and the core domains of this transactivator were necessary and sufficient to activate transcription when TBP was overexpressed. In vitro experiments showed that Tat binds specifically to TBP, and follow up in vivo experiments indicated a correlation between the ability of different Tat mutants to bind TBP and their capacity to activate transcription in vivo [27,49-51]. Still other studies that looked at the interaction of TBP and Tat concluded that activation of the LTR requires steps in addition to TBP recruitment [52].
The Hernandez lab has previously shown that TBP bound to the TATA box was required for the synthesis of short and full-length transcripts as well as for Tat activation and that both yeast TBP and the carboxy-terminal domain of human TBP could replace full-length human TBP for these processes [53]. Similar studies from the Lania lab indicated similar activation by a TBP fusion. For instance, to determine the synergy between Tat and GAL4-TBP in the absence of any DNA-bound activator, the G1-38HIV reporter was transfected into HeLa cells with the GAL4-hTBP and a Tat expression vector. Tat alone had no effect on transcription, however, co-expression of Tat strongly stimulated GAL4-hTBP transcription in the absence of any DNA-bound activator. Synergy between Tat and DNA-bound TBP protein was further confirmed by analysis of the levels of specific transcripts, which were determined by RNase protection assay [54]. Therefore, artificial recruitment of human TBP to the enhancerless HIV minimal promoter was found to trigger gene expression, and coexpression of Tat resulted in a marked synergy. The functional cooperation between TBP and Tat was further demonstrated using the Drosophila Schneider SL2 cells [55].
Finally, with regard to the functional significance of Tat's role in transcription complex assembly and levels of nonprocessive transcription from the HIV-1 LTR, two manuscripts from the Jeang lab are worth noting. First, a central question was asked in whether the LTR promoter "presynthesizes" short nascent TAR RNA-containing transcripts that remain poised awaiting Tat. To address this question, the investigators used a self-cleaving ribozyme to define a time window during which Tat action occurred, which measured Tat trans-activation against two biological processes: RNA chain elongation and RNA self-cleavage [56]. To do this, they placed a rapidly self-cleaving ribozyme downstream of TAR. The experimental model assumed that if the ribozyme self-cleavage reaction was sufficiently rapid then it should sever the TAR-Tat complex that was attached to the nascent RNA chain and thus prevent an interaction with the LTR promoter. Therefore, the speed of one process (trans-activation) was compared against another (RNA chain elongation leading to self-cleavage). From their experiments, they concluded that an accumulation of paused TAR transcripts between +42 and +80 was unlikely, and the evidence that rapid cleavage at +80 did affect (rate determine) the overall trans-activation process was not compatible with pausing at this location on the DNA template. Control experiments demonstrated that the observed reduction in expression was specific for a functional ribozyme and specific for trans-activation (as opposed to a perturbation in basal activity or in RNA stability).
Second, when examining the short (S) and long (L) form of HIV-1 RNA in an integrated provirus setting in vivo, they suggested that S RNAs, while seen in unintegrated DNA and/or cell-free assays, were not prevalent in the context of integrated proviruses [57]. Basal transcription from a vector containing SV40 sequence (pHIVCATSV) in Cos cells was characterized by an abundance of S transcripts, while a normal HIV vector (pLTRCAT) produced no such RNAs. With Tat, both plasmids transcribed comparable amounts of L transcripts. They concluded that abortive transcripts may simply reflect transcription that occurs as a consequence of replication induced by T antigen in cell lines tested. These data were also consistent with earlier reports on Tat's effect of TC complex stability when using cell-free assays [48].
While focus of the studies on Tat function was heavily placed on the role of cellular kinases, protein phosphatases might also play an important role in the early stages of HIV-1 transcription. Ammosova et al. [58] have shown that PP1 and PP2A dephosphorylate CDK9 and that inhibition of PP1 or PP2A phosphatase activity decreases HIV transcription in vitro and in vivo. While the authors concentrated on the activity of PP1 and PP2A on autophosphorylation sites, which include the activation site at Thr 186 and more C-terminal phosphorylation sites, it is possible that these phosphatases play a role in removing inhibitory CDK9 phosphorylation sites in the preinitiation complex [[23], M. Zhou, personal communication].
Chromatin structure
In considering the effect of Tat on transcription initiation and elongation, the effect of chromatin structure on the integrated genome must be considered. Investigators have shown that chromatin exerts a strong repressive role on transcription initiation. Interestingly, in a 2003 study using chronically infected U1 cells treated with phorbol ester, Lusic et al. reported that Tat promotes the specific recruitment of histone acetyltransferases to the viral promoter, facilitating acetylation of histones H3 and H4 at distinct nucleosomal regions, before the onset of viral mRNA transcription [59]. In a separate study, Kiefer et al. reported that nucleosome remodeling, not histone acetylation, is the limiting step in transcriptional activation in U1 promonocytes [60]. It is possible therefore, that Tat facilitates chromatin modifications, assembly of the initiation complex and transcription elongation in a series of sequential, coordinated events that leads to high levels of HIV transcription.
Similarities to viral transactivators Herpesvirus VP16 and IE, Adenovirus E1A and SV40 T-antigen and HTLV-1 Tax
We should not be surprised by the complexity of the Tat transactivation process and the multifaceted effect of Tat on multiple transcription factors involved in LTR regulation. Examination of viral activators and their mechanism of activation indicate how small DNA or RNA viruses have evolved intricate mechanisms for controlling viral and cellular gene expression. For example, the Herpesvirus VP16 activation domain can be divided into two modules – an N-terminal subdomain (VPN) and a C-terminal subdomain (VPC). It has been shown [61] that VPC stimulates core promoters that are either independent of or dependent on TAFs (TATA box Binding Protein-Associated Factors). In contrast, VPN only activates the TAF-independent core promoter, and this activity increases in a synergistic fashion when VPN is dimerized (VPN2). The VPN subdomain of VP16 also facilitates assembly of a transcriptional complex containing TBP: TFIIA:TFIIB, which lacks TAFs, and provides a mechanism that could function at TAF-independent promoters. Thus, the viral activator facilitates transcription through multiple functional pathways.
Along the same lines, biochemical and genetic evidence has suggested that the Herpesvirus IE proteins may perform functions similar to those of the TAFs in the transcriptional complex. The IE proteins expressed from the intact major IE gene, and to a lesser extent IEP86 alone, could rescue the temperature-sensitive (ts) transcriptional defect in TAFII250 BHK-21 ts13 cell line [62].
The adenovirus E1A protein is also a well studied transcription activator. The 48-amino-acid conserved region 3 (CR3) of E1A, which is responsible for mediating transactivation, appears to target several proteins of the transcription initiation complex, including ATF-2, and components of the basal transcription factor TFIID, including TBP, hTAFII250, hTAFII55, and hTAFII135 [63]. This interaction allows E1A to stabilize the TFIID-TFIIA complex to increase the level of activated transcription in vivo.
Another viral activator, SV40 large T-antigen has also been shown to specifically enhance the formation of the TBP-TFIIA complex on the TATA element. The ability to facilitate TBP/TFIIA binding was complex and promoter dependent since T-antigen could activate simple promoters containing the TATA elements from the hsp70 and c-fos gene promoters but failed to significantly activate similar promoters containing the TATA elements from the promoters of the SV40 early and adenovirus E2a genes. Furthermore, the ability to stabilize the TBP-TFIIA complex on the hsp70 and c-fos TATA elements, and not on the SV40 early and E2A TATA elements, correlated with the ability or inability to activate promoters containing these TATA elements [64]. Interestingly, in the ts13 cell line, T-antigen could rescue the temperature-sensitive (ts) defect in TAFII250. In contrast, neither E1A, small t-antigen, nor mutants of T-antigen defective in transcriptional activation were able to rescue the ts defect [65], further implying that T-antigen may act like a TAF activator.
Finally, while investigating the effect of HTLV-1 Tax on the pre-initiation complex assembly, it has been shown that Tax facilitates the binding of a variety of transcription factors including CREB, TFIIA, CBP/p300 and PCAF [66-69]. Interestingly, Caron et al., have shown that transactivation by Tax was correlated with its ability to interact with the C-terminal moiety of the TBP and hTAFII28 in transfected HeLa cells [70]. An increase in the intracellular concentration of hTAFII28 augmented transactivation by Tax. This effect was also seen in COS-7 cells that have low levels of endogenous TAFII28. TBP and hTAFII28 also cooperated to allow Tax activation of the entire HTLV-I promoter and to partially rescue the phenotype of Tax mutants that had an impaired ability to activate transcription. The authors speculated that since TBP was present in all three cellular RNA polymerases, an increase in the concentration of hTAFII28, which binds directly to TBP, may compete with the TAFIs and TAFIIIs and drive more TBP into the formation of a TFIID complex interacting with Tax. According to this model, overexpression of both TBP and hTAFII28 would most efficiently raise the concentration of TFIID complexes capable of functioning with Tax.
Future considerations
Recent technical advances, such as ChIP and siRNA assays, allowed Raha et al. [41] to more clearly demonstrate that Tat facilitates TC assembly at the HIV initiation site in vivo. The results of this study are consistent with and supported by previous studies which demonstrated that Tat, pTEFb and HATs are present on the HIV promoter and support a role of Tat in transcriptional initiation [23,32,59,71]. It should be noted that, in addition to technical advances, the ability to detect Tat in transcription initiation is likely dependent upon the experimental system. Along these lines, it should be noted that other recent ChIP data are more consistent with an effect of Tat at transcription elongation. Bres et al. recently reported that in HeLa P4 cells SPT5, SPT6, RNAP II and Ser 5-P CTD were present on the integrated HIV promoter in the absence of Tat and ongoing transcription [72]. Future work will continue on the exciting and multifaceted and perhap sequential role of Tat in chromatin remodeling, preinitiation complex assembly, elongation, and processing and will include questions such as: 1) How P-TEFb, which was initially discovered as an elongation factor, selectively recruits TBP alone or in complex with other TAFs and activators; 2) Are there TBP associated complexes that are selective to Tat and not to other cellular promoters, and can they be purified to homogeneity; 3) Is the TBP recruited from the PolI or PolIII TC; 4) Does Tat act similarly to TAF subunits replacing some or all of the TFIID TAF subunits in vivo; 5) Can the data be reproduced in primary T- and monocytic latent patient samples? One thing is certain however. More research and funding is needed to define various mechanisms of Tat function in the hope that the data will result in finding the very first specific HIV transcription inhibitor in vivo.
Figure 1 The HIV promoter is comprised of a series of transcription control elements including NF-kB, Sp1, TATA box, RNA initiation site and the downstream TAR RNA enhancer element. In the presence of Tat, a complex interaction between activators which include NF-kB and/or Sp1 bind to the upstream control region and interact with transcription factors which include, but may not be limited to, TBP, TFIIH, P-TEFb and RNA Pol II. Data from several laboratories now support a role for Tat in transcription complex assembly. Tat and P-TEFb facilitate the binding of TBP to the complex, setting the stage for binding of other basal transcription factors and assembly of the preinitiation complex. In the initiation complex, although both TFIIH and P-TEFb are present, the Pol II CTD is phosphorylated primarily by TFIIH at Ser5 (black). Following synthesis of the TAR RNA enhancer and loss of TFIIH from the elongation complex, P-TEFb is autophosphorylated at Thr186. Transcription elongation requires the interaction of Tat and P-TEFb with the TAR RNA which facilitates the phosphorylation of the Pol II CTD at Ser2 (red) and Ser5 (yellow), as well as the phosphorylation of Tat cofactors Tat-SF1 and SPT5. Whether the Tat and P-TEFb bound to TAR are transferred from the initiation complex, or represent the binding of additional Tat and P-TEFb remains to be established. The Tat-modified kinase activity of P-TEFb is preferentially sensitive to low concentrations of DRB or flavopiridol. This model assumes that the Tat and P-TEFb associated with the initiation complex transfers to the TAR RNA enhancer and perhaps to the elongation complex, a point that has not yet been demonstrated.
Acknowledgements
The authors would like to thank the Brady and Kashanchi lab members for their helpful and critical comments, Ms. Lynne Mied and Cynthia de la Fuente for assisting with the manuscript and Dr. Sergei Nekhai for his contribution on the phosphatase section. This work was supported by grants from the George Washington University REF fund (A. Vertes and F. Kashanchi) and NIH grants AI44357, AI43894, and 13969 to F.K.
==== Refs
Pumfery A Deng L Maddukuri A de la Fuente C Li H Wade JD Lambert L Kumar A Kashanchi F Chromatin Remodeling and Modification during HIV-1 Tat-activated Transcription Curr HIV Res 2003 1 261 2741 15046251 10.2174/1570162033485186
Berkhout B Gatignol A Rabson AB Jeang KT TAR-independent activation of the HIV-1 LTR: evidence that tat requires specific regions of the promoter Cell 1990 62 757 767 2201451 10.1016/0092-8674(90)90120-4
Calnan BJ Biancalana S Hudson D Frankel AD Analysis of arginine-rich peptides from the HIV Tat protein reveals unusual features of RNA-protein recognition Genes Dev 1991 5 201 210 1899841
Cato AC Henderson D Ponta H The hormone response element of the mouse mammary tumour virus DNA mediates the progestin and androgen induction of transcription in the proviral long terminal repeat region Embo J 1987 6 363 368 3034586
Selby MJ Bain ES Luciw PA Peterlin BM Structure, sequence, and position of the stem-loop in tar determine transcriptional elongation by tat through the HIV-1 long terminal repeat Genes Dev 1989 3 547 58 2470647
Kao SY Calman AF Luciw PA Peterlin BM Anti-termination of transcription within the long terminal repeat of HIV-1 by tat gene product Nature 1987 330 489 493 2825027 10.1038/330489a0
Laspia MF Rice AP Mathews MB HIV-1 Tat protein increases transcriptional initiation and stabilizes elongation Cell 1989 59 283 92 2553266 10.1016/0092-8674(89)90290-0
Garcia JA Harrich D Soultanakis E Wu F Mitsuyasu R Gaynor RB Human immunodeficiency virus type 1 LTR TATA and TAR region sequences required for transcriptional regulation Embo J 1989 8 765 78 2721501
Dingwall C Ernberg I Gait MJ Green SM Heaphy S Karn J Lowe AD Singh M Skinner MA Valerio R Human immunodeficiency virus 1 tat protein binds trans-activation-responsive region (TAR) RNA in vitro Proc Natl Acad Sci U S A 1989 86 6925 9 2476805
Marshall NF Price DH Control of formation of two distinct classes of RNA polymerase II elongation complexes Mol Cell Biol 1992 12 2078 2090 1569941
Jang MK Mochizuki K Zhou M Jeong HS Brady JN Ozato K The bromodomain protein Brd4 is a positive regulatory component of P-TEFb and stimulates RNA polymerase II-dependent transcription Mol Cell 2005 19 523 34 16109376 10.1016/j.molcel.2005.06.027
Fujinaga K Cujec TP Peng J Garriga J Price DH Grana X Peterlin BM The ability of positive transcription elongation factor B to transactivate human immunodeficiency virus transcription depends on a functional kinase domain, cyclin T1, and Tat J Virol 1998 72 7154 7159 9696809
Garber ME Wei P KewalRamani VN Mayall TP Herrmann CH Rice AP Littman DR Jones KA The interaction between HIV-1 Tat and human cyclin T1 requires zinc and a critical cysteine residue that is not conserved in the murine CycT1 protein Genes Dev 1998 12 3512 3527 9832504
Isel C Karn J Direct evidence that HIV-1 Tat stimulates RNA polymerase II carboxyl-terminal domain hyperphosphorylation during transcriptional elongation J Mol Biol 1999 290 929 941 10438593 10.1006/jmbi.1999.2933
Jones KA Taking a new TAK on tat transactivation Genes Dev 1997 11 2593 2599 9334323
Ramanathan Y Reza SM Young TM Mathews MB Pe'ery T Human and rodent transcription elongation factor P-TEFb: interactions with human immunodeficiency virus type 1 tat and carboxy-terminal domain substrate J Virol 1999 73 5448 5458 10364292
Herrmann CH Rice AP Lentivirus Tat proteins specifically associate with a cellular protein kinase, TAK, that hyperphosphorylates the carboxyl-terminal domain of the large subunit of RNA polymerase II: candidate for a Tat cofactor J Virol 1995 69 1612 1620 7853496
Nekhai S Shukla RR Kumar A A human primary T-lymphocyte-derived human immunodeficiency virus type 1 Tat-associated kinase phosphorylates the C-terminal domain of RNA polymerase II and induces CAK activity J Virol 1997 71 7436 7441 9311822
Zhou M Halanski MA Radonovich MF Kashanchi F Peng J Price DH Brady JN Tat modifies the activity of CDK9 to phosphorylate serine 5 of the RNA polymerase II carboxyl-terminal domain during human immunodeficiency virus type 1 transcription Mol and Cell Biol 2000 20 5077 5086 10866664 10.1128/MCB.20.14.5077-5086.2000
Yang XJ Ogryzko VV Nishikawa J Howard BH Nakatani Y A p300/CBP-associated factor that competes with the adenoviral oncoprotein E1A Nature 1996 382 319 324 8684459 10.1038/382319a0
Garcia-Martinez LF Mavankal G Neveu JM Lane WS Ivanov D Gaynor RB Purification of a Tat-associated kinase reveals a TFIIH complex that modulates HIV-1 transcription Embo J 1997 16 2836 2850 9184228 10.1093/emboj/16.10.2836
Parada CA Roeder RG Enhanced processivity of RNA polymerase II triggered by Tat-induced phosphorylation of its carboxy-terminal domain Nature 1996 384 375 378 8934526 10.1038/384375a0
Zhou M Deng L Lacoste V Park HU Pumfery A Kashanchi F Brady JN Kumar A Coordination of transcription factor phosphorylation and histone methylation by the P-TEFb kinase during human immunodeficiency virus type 1 transcription J Virol 2004 78 13522 33 15564463 10.1128/JVI.78.24.13522-13533.2004
Jeang KT Chun R Lin NH Gatignol A Glabe CG Fan H In vitro and in vivo binding of human immunodeficiency virus type 1 Tat protein and Sp1 transcription factor J Virol 1993 67 6224 6233 7690421
Kashanchi F Duvall JF Dittmer J Mireskandari A Reid RL Gitlin SD Brady JN Involvement of transcription factor YB-1 in human T-cell lymphotropic virus type I basal gene expression J Virol 1994 68 561 565 8254772
Majello B Napolitano G Lania L Recruitment of the TATA-binding protein to the HIV-1 promoter is a limiting step for Tat transactivation AIDS 1998 12 1957 1964 9814863
Veschambre P Simard P Jalinot P Evidence for functional interaction between the HIV-1 Tat transactivator and the TATA box binding protein in vivo J Mol Biol 1995 250 169 180 7608968 10.1006/jmbi.1995.0368
Deng L Ammosova T Pumfery A Kashanchi F Nekhai S HIV-1 Tat interaction with RNA polymerase II C-terminal domain (CTD) and a dynamic association with CDK2 induce CTD phosphorylation and transcription from HIV-1 promoter J Biol Chem 2002 277 33922 33929 12114499 10.1074/jbc.M111349200
Kamine J Elangovan B Subramanian T Coleman D Chinnadurai G Identification of a cellular protein that specifically interacts with the essential cysteine region of the HIV-1 Tat transactivator Virology 1996 216 357 366 8607265 10.1006/viro.1996.0071
Cujec TP Cho H Maldonado E Meyer J Reinberg D Peterlin BM The human immunodeficiency virus transactivator Tat interacts with the RNA polymerase II holoenzyme Mol and Cell Biol 1997 17 1817 1823 9121429
Mavankal G Ignatius Ou SH Oliver H Sigman D Gaynor RB The human immunodeficiency virus transactivator Tat interacts with the RNA polymerase II holoenzyme Proc Natl Acad Sci U S A 1996 93 2089 2094 8700889 10.1073/pnas.93.5.2089
Marzio G Tyagi M Gutierrez MI Giacca M HIV-1 tat transactivator recruits p300 and CREB-binding protein histone acetyltransferases to the viral promoter Proc Natl Acad Sci U S A 1998 95 13519 13524 9811832 10.1073/pnas.95.23.13519
Benkirane M Chun RF Xiao H Ogryzko VV Howard BH Nakatani Y Jeang KT Activation of integrated provirus requires histone acetyltransferase. p300 and P/CAF are coactivators for HIV-1 Tat J Biol Chem 1998 273 24898 24905 9733796 10.1074/jbc.273.38.24898
Col E Caron C Seigneurin-Berny D Gracia J Favier A Khochbin S The histone acetyltransferase, hGCN5, interacts with and acetylates the HIV transactivator, Tat J Biol Chem 2001 276 28179 28184 11384967 10.1074/jbc.M101385200
Garber ME Jones KA HIV-1 Tat: coping with negative elongation factors Curr Opin Immunol 1999 11 460 465 10448148 10.1016/S0952-7915(99)80077-6
Karn J Tackling Tat J Mol Biol 1999 293 235 254 10550206 10.1006/jmbi.1999.3060
Brigati C Giacca M Noonan DM Albini A HIV Tat, its TARgets and the control of viral gene expression FEMS Microbiol Lett 2003 220 57 65 12644228 10.1016/S0378-1097(03)00067-3
Giacca M The HIV-1 Tat protein: a multifaceted target for novel therapeutic opportunities Curr Drug Targets Immune Endocr Metabol Disord 2004 4 277 285 15578978 10.2174/1568008043339767
Bannwarth S Gatignol A HIV-1 TAR RNA: the target of molecular interactions between the virus and its host Curr HIV Res 2005 3 61 71 15638724 10.2174/1570162052772924
Barboric M Peterlin BM A new paradigm in eukaryotic biology: HIV Tat and the control of transcriptional elongation PLoS Biol 2005 3 e76 15719065 10.1371/journal.pbio.0030076
Raha T Cheng SW Green MR HIV-1 Tat stimulates transcription complex assembly through recruitment of TBP in the absence of TAFs PLoS Biol 2005 3 e44 15719058 10.1371/journal.pbio.0030044
Brown JA Howcroft TK Singer DS HIV Tat protein requirements for transactivation and repression of transcription are separable J Acquir Immune Defic Syndr Hum Retrovirol 1998 17 9 16 9436753
Howcroft TK Palmer LA Brown J Rellahan B Kashanchi F Brady JN Singer DS HIV Tat represses transcription through Sp1-like elements in the basal promoter Immunity 1995 3 127 138 7621073 10.1016/1074-7613(95)90165-5
Carroll IR Wang J Howcroft TK Singer DS HIV Tat represses transcription of the beta 2-microglobulin promoter Mol Immunol 1998 35 1171 1178 10199391 10.1016/S0161-5890(98)00107-2
Weissman JD Brown JA Howcroft TK Hwang J Chawla A Roche PA Schiltz L Nakatani Y Singer DS HIV-1 tat binds TAFII250 and represses TAFII250-dependent transcription of major histocompatibility class I genes Proc Natl Acad Sci U S A 1998 95 11601 11606 9751712 10.1073/pnas.95.20.11601
Kashanchi F Piras G Radonovich MF Duvall JF Fattaey A Chiang CM Roeder RG Brady JN Direct interaction of human TFIID with the HIV-1 transactivator tat Nature 1994 367 95 9 10.1038/367295a0
Chiang CM Roeder RG Cloning of an intrinsic human TFIID subunit that interacts with multiple transcriptional activators Science 1995 267 531 6 7824954
Bohan CA Kashanchi F Ensoli B Buonaguro L Boris-Lawrie KA Brady JN Analysis of Tat transactivation of human immunodeficiency virus transcription in vitro Gene Expr 1992 2 391 407 1282057
Wang Z Morris GF Rice AP Xiong W Morris CB Wild-type and transactivation-defective mutants of human immunodeficiency virus type 1 Tat protein bind human TATA-binding protein in vitro J Acquir Immune Defic Syndr Human Retrovirol 1996 12 128 38 8680883
Kashanchi F Khleif SN Duvall JF Sadaie MR Radonovich MF Cho M Martin MA Chen SY Weinmann R Brady JN Interaction of human immunodeficiency virus type 1 Tat with a unique site of TFIID inhibits negative cofactor Dr1 and stabilizes the TFIID-TFIIA complex J Virol 1996 70 5503 10 8764062
Veschambre P Roisin A Jalinot P Biochemical and functional interaction of the human immunodeficiency virus type 1 Tat transactivator with the general transcription factor TFIIB J Gen Virol 1997 78 2235 45 9292011
Xiao H Lis JT Jeang KT Promoter activity of Tat at steps subsequent to TATA-binding protein recruitment Mol Cell Biol 1997 17 6898 6905 9372921
Pendergrast PS Morrison D Tansey WP Hernandez N Mutations in the carboxy-terminal domain of TBP affect the synthesis of human immunodeficiency virus type 1 full-length and short transcripts similarly J Virol 1996 70 5025 34 8764009
Majello B Napolitano G De Luca P Lania L Recruitment of human TBP selectively activates RNA polymerase II TATA-dependent promoters J Biol Chem 1998 273 16509 16 9632719 10.1074/jbc.273.26.16509
Majello B Napolitano G Lania L Recruitment of the TATA-binding protein to the HIV-1 promoter is a limiting step for Tat transactivation AIDS 1998 12 1957 64 9814863
Jeang KT Berkhout B Kinetics of HIV-1 long terminal repeat transactivation. Use of intragenic ribozyme to assess rate-limiting steps J Biol Chem 1992 267 17891 9 1517225
Jeang KT Berkhout B Dropulic B Effects of integration and replication on transcription of the HIV-1 long terminal repeat J Biol Chem 1993 268 24940 9 8227056
Ammosova T Washington K Debebe Z Brady J Nekhai S Dephosphorylation of CDK9 by protein phosphatase 2A and protein phosphatase-1 in Tat-activated HIV-1 transcription Retrovirology 2005 2 47 16048649 10.1186/1742-4690-2-47
Lusic M Marcello A Cereseto A Giacca M Regulation of HIV-1 gene expression by histone acetylation and factor recruitment at the LTR promoter Embo J 2003 22 6550 6561 14657027 10.1093/emboj/cdg631
Kiefer HL Hanley TM Marcello JE Karthik AG Viglianti GA Retinoic acid inhibition of chromatin remodeling at the human immunodeficiency virus type 1 promoter. Uncoupling of histone acetylation and chromatin remodeling J Biol Chem 2004 279 43604 43613 15299018 10.1074/jbc.M408069200
Hori RoderickT Shuping Xu Xianyuan Hu Sung Pyo TFIIB-facilitated recruitment of preinitiation complexes by a TAF-independent mechanism Nucleic Acids Res 2004 32 13 10.1093/nar/gkh711
Lukac DM Harel NY Tanese N Alwine JC TAF-Like Functions of Human Cytomegalovirus Immediate-Early Proteins J Virol 1997 71 7227 7239 9311796
Mazzarelli JM Mengus G Davidson I Ricciardi RP The Transactivation Domain of Adenovirus E1A Interacts with the C Terminus of Human TAFII135 J Virol 1997 71 7978 7983 9311891
Damania B Lieberman P Alwine JC Simian Virus 40 Large T Antigen Stabilizes the TATA-Binding Protein-TFIIA Complex on the TATA Element Mol and Cell Biol 1998 17 3926 3935 9632777
Damania B Alwine JC TAF-like function of SV40 large T antigen Genes Dev 1996 10 1369 1381 8647434
Andrisani O Dixon JE Identification and purification of a novel 120-kDa protein that recognizes the cAMP-responsive element J Biol Chem 1990 265 3212 8 2137455
Clemens KE Piras G Radonovich MF Choi KS Duvall JF DeJong J Roeder R Brady JN Interaction of the human T-cell lymphotropic virus type 1 tax transactivator with transcription factor IIA Mol Cell Biol 1996 16 4656 64 8756622
Kwok RP Laurance ME Lundblad JR Goldman PS Shih H Connor LM Marriott SJ Goodman RH Control of cAMP-regulated enhancers by the viral transactivator Tax through CREB and the co-activator CBP Nature 1996 380 642 6 8602268 10.1038/380642a0
Jiang H Lu H Schiltz RL Pise-Masison CA Ogryzko VV Nakatani Y Brady JN PCAF interacts with tax and stimulates tax transactivation in a histone acetyltransferase-independent manner Mol Cell Biol 1999 19 8136 45 10567539
Caron CC Mengus G Dubrowskaya V Roison A Davidson I Jalinot P Human TAFII28 interacts with the human T cell leukemia virus type I Tax transactivator and promotes its transcriptional activity Proc Natl Acad Sci U S A 1997 94 3662 3667 9108034 10.1073/pnas.94.8.3662
Ping YH Rana TM Tat-associated kinase (P-TEFb): a component of transcription preinitiation and elongation complexes J Biol Chem 1999 274 7399 7404 10066804 10.1074/jbc.274.11.7399
Bres V Gomes N Pickle L Jones KA A human splicing factor, SKIP, associates with P-TEFb and enhances transcription elongation by HIV-1 Tat Genes Dev 2005 19 1211 1226 15905409 10.1101/gad.1291705
|
16280076
|
PMC1308864
|
CC BY
|
2021-01-04 16:36:41
|
no
|
Retrovirology. 2005 Nov 9; 2:69
|
utf-8
|
Retrovirology
| 2,005 |
10.1186/1742-4690-2-69
|
oa_comm
|
==== Front
RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-701628588510.1186/1742-4690-2-70ResearchPresence of a functional but dispensable Nuclear Export Signal in the HTLV-2 Tax protein Chevalier Sébastien A [email protected] Laurent [email protected] Sara [email protected] Antoine [email protected] Lars [email protected] Renaud [email protected] Unité d'Epidémiologie et Physiopathologie des Virus Oncogènes, Institut Pasteur, Paris, France2 Center for Biological Sequence Analysis, BioCentrum-DTU, The Technical University of Denmark, Building 208 DK-2800, Lyngby, Denmark2005 14 11 2005 2 70 70 14 10 2005 14 11 2005 Copyright © 2005 Chevalier et al; licensee BioMed Central Ltd.2005Chevalier et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Human T-cell leukemia virus type 1 and type 2 are related human retroviruses. HTLV-1 is the etiological agent of the Adult T-cell Leukemia/Lymphoma and of the Tropical Spastic Paraparesis/HTLV-1 Associated Myelopathy, whereas, HTLV-2 infection has not been formally associated with any T-cell malignancy. HTLV-1 and 2 genomes encode, respectively, the Tax1 and Tax2 proteins whose role is to transactivate the viral promoter. HTLV-1 and HTLV-2 Tax sequences display 28% divergence at the amino acid level. Tax1 is a shuttling protein that possesses both a non canonical nuclear import (NLS) and a nuclear export (NES) signal. We have recently demonstrated that Tax1 and Tax2 display different subcellular localization and that residues 90–100 are critical for this process. We investigate in the present report, whether Tax2 also possesses a functional NES.
Results
We first used a NES prediction method to determine whether the Tax2 protein might contain a NES and the results do suggest the presence of a NES sequence in Tax2. Using Green Fluorescent Protein-NES (GFP-NES) fusion proteins, we demonstrate that the Tax2 sequence encompasses a functional NES (NES2). As shown by microscope imaging, NES2 is able to mediate translocation of GFP from the nucleus, without the context of a full length Tax protein. Furthermore, point mutations or leptomycin B treatment abrogate NES2 function. However, within the context of full length Tax2, similar point mutations in the NES2 leucine rich stretch do not modify Tax2 localization. Finally, we also show that Tax1 NES function is dependent upon the positioning of the nuclear export signal "vis-à-vis" GFP.
Conclusion
HTLV-2 Tax NES is functional but dispensable for the protein localization in vitro.
==== Body
Background
HTLV-1 and HTLV-2 are closely related retroviruses that infect T-cells in vivo, with a probable preferential tropism for CD4+ and CD8+ cells respectively [1]. HTLV-1 is the etiological agent of the Adult T-cell Leukemia/Lymphoma (ATLL) and of the Tropical Spastic Paraparesis/HTLV-1 Associated Myelopathy (TSP/HAM), while HTLV-2 infection, even if originally described in a patient suffering of atypical hairy T-cell leukemia, has only been linked to infrequent cases of TSP/HAM "like" disease [2-4]. Both HTLV-1 and HTLV-2 genomes encode a viral transactivator (Tax1 and Tax2 respectively). Tax1 has an oncogenic potential and is responsible for cell-transformation in vitro [5,6]. Tax1 and Tax2 display approximately 75% nucleotide sequence homology. Strikingly however, several reports have now demonstrated that although the critical functional regions of the proteins are well conserved (i.e. NF-κB and CREB/ATF activation domains), the two transactivators exhibit a number of major phenotypical differences [1,7-18]. Nevertheless, Tax2 is capable of immortalizing human lymphocytes and, although to a lesser extent than Tax1, of transforming rat cells in vitro [10,19].
Eukaryotic cells are compartmentalized into the cytoplasm and the nucleus by the nuclear envelope [20,21]. The nuclear envelope contains nuclear pore complexes (NPCs), which mediate the traffic of molecules between the two compartments. The nucleo-cytoplasmic traffic of large molecules is regulated by specific nuclear import and export systems. Proteins that contain classical Nuclear Localization Signals (NLSs) are imported into the nucleus by importin α/β protein heterodimers. So far, six importin α family members and one importin β have been described [22]. Importin α binds to NLS containing proteins, whereupon importin β is responsible for the docking of the importin cargo complex to the cytoplasmic side of the NPC, followed by translocation of the complex through the NPC. A classical monopartite NLS consists of a stretch of basic amino acids such as arginines and lysines. Contrary to this, the Nuclear Export Signal (NES) generally consists of a leucine/isoleucine-rich sequence [23]. The classical NES pattern is L-x(2,3)- [LIVFM]-x(2,3)-L-x- [LI], where L can either be L, I, V, F or M, but many known NES regions do not conform to these limitations [24]. For example, the spacing between the hydrophobic residues is variable and NES regions can also be rich in glutamate, aspartate and serine [23]. The first nuclear export pathway to be discovered involved the chromosome region maintenance 1 (CRM1) receptor, exporting proteins containing a nuclear export signal (NES) [25]. CRM1 binds to a Nuclear Export Signal (NES)-containing protein and to the NPC. Several ways of regulating NES-dependent export have been reported, including masking or unmasking the NES and post-translational modifications of the NES-containing protein [26].
Cellular fractionation and immunofluorescence experiments performed with HTLV-1 infected and Tax1 transfected cells have demonstrated that Tax1 was present both in the nuclear and cytoplasmic fractions. However, the distribution of the protein between these two compartments is unequal and depends on the cell-line used [27-31]. The Tax1 48 amino terminal sequence contains a non-canonical functional NLS [32] that allows the protein to enter the nucleus, where Tax1 localizes to discrete nuclear bodies (also called Tax Speckled Structures (TSS) [33]. In addition, Tax1 also contains a "Rev-like" Nuclear Export Signal (NES) spanning from amino acid 189 to 202. This NES is insensitive to leptomycin B within the context of the full-length protein [27]. Both localization signals (NLS and NES) are likely to be involved in the shuttling of Tax1, but this process is still not clearly understood [34].
We have recently reported that, although Tax2 contains a functional NLS domain, the protein localizes predominantly to the cytoplasm in HTLV-2 immortalized or transformed infected T-cells as well as in Tax2 transfected cells [16]. These results were further confirmed in another laboratory [35] which also demonstrated that the NLS domain was confined to the 40 first N-terminal amino acids. We also demonstrated that the region spanning amino acids 90 to 100 was critical for Tax2 localization [16]. The recent report of a Tax1 NES sequence prompted us to examine the possible presence of a NES in Tax2. In addition to the 90–100 domain, this sequence could serve as a second domain involved in Tax2 localization. We show in this report that, although HTLV-2 Tax protein contains a NES sequence that is active without the context of a full-length protein, this domain is dispensable for the Tax2 localization.
Results
HTLV-2 Tax protein sequence contains a putative NES domain
We lately demonstrated that the HTLV-2 Tax protein has an intracellular localization that is different from that of Tax1, both in infected and transfected cells (i.e. Tax2 localizes more to the cytoplasm than Tax1) and that, within the Tax sequence, the 90–100 domain was critical for the protein localization [16]. These results were confirmed lately [35]. Another recent article reported that, in addition to the previously characterized NLS, HTLV-1 Tax protein also contains a Nuclear Export Signal (NES) comprising amino acids 189 to 202 (KRIEELLYKISLTT). This sequence contains a string of hydrophobic amino acids (I191, L195, I198 and L200) [27] and has the ability to redirect the Green Fluorescent Protein (GFP) to the cytoplasm. Within the Tax1 NES sequence, residues L195 (formerly named L194 [27]) and L200 appear to be critical for the Tax1 NES function. As an example, when Tax1 L200 is mutated to an alanine, the GFP-Tax1 localization is altered [27].
In order to identify whether Tax2 contains a similar sequence, we first used the NetNES prediction method [23]. (Figure 1). From this analysis, residue L188 of Tax2 was predicted to be part of the Tax2 NES domain. Interestingly, L188 is absent from the sequence of Tax1, where the position is occupied by a tyrosine (data not shown). The amino acid comparison of Tax1 and Tax2 also reveals that the two sequences are very similar in the 189 to 202 amino acids region, with observed differences at positions 191, 198, 199 and 201 (Figure 2A).
Figure 1 Output after submission of the complete Tax2 amino acid sequence to the NetNES website [23]. Output consists of two parts; one is a listing of all residues with the individual scores noted (see column heading), the other is a graphical plot of the values given in the table. The prediction server calculates the NES score from the HMM and Artificial Neural Network (ANN) scores but all three values are given for each residue. If the calculated 'NES score' exceeds the threshold, then that particular residue is expected to participate in a nuclear export signal. This is denoted with a 'Yes' in the column "Predicted".
Figure 2 Without the context of the full length protein, NES2 can redirect GFP to the cytoplasm, while NES1 function depends on its positioning vis-à-vis GFP. (A): Sequence alignment of a consensus NES sequence with Tax1 NES and Tax2 putative NES. (B): HeLa cells were transiently transfected with NES-EGFP and GFP-NES plasmids. Twenty-four hours after transfection, the cells were washed with PBS, fixed with 4% paraformaldehyde, mounted with DAPI-containing mounting medium and visualized with a Zeiss Axioplan 2 imaging microscope X40 using a Zeiss Axiocam HRc (color) camera and the Zeiss Apotome software. Images of cells that are representative of the entire population are shown. (C): Western-blot analysis of cytoplasmic and nuclear cell fractions. 293T cell fractions were subjected to electrophoresis on a 10 % TG gel and probed with a GFP antibody. The western-blot results are representative of four independent experiments.
We set out to investigate whether, despite these differences, the Tax2 putative NES was functional. To this end, we affixed the 189–202 amino-acid domain of Tax2 to the N-terminus of the GFP sequence (NES2-EGFP) using the pEGFP-N1 vector as previously reported [27]. The NES2-EGFP construct was then transiently transfected in 293T (data not shown) and Hela cells, as these cells have frequently been used for Tax localization studies [17,27]. As a positive control, the NES Tax1 sequence was also fused to the N-terminus of the GFP (NES1-EGFP). In the absence of a Tax NES sequence, the GFP protein is nearly equally distributed between the cytoplasm and the nucleus of the transfected cells ([16] and data not shown). However, the GFP signal was almost entirely cytoplasmic when the protein was fused to the Tax2 putative NES (Figure 2B panel a and Figure 2C for fractionation). This suggests that this latter sequence mediates an active transport of GFP in vitro. Unexpectedly, and contrary to a previous report [27], the NES1-EGFP fusion protein was diffused in both the nucleus and the cytoplasm with a nuclear content that was much higher than that of NES2-EGFP (Figure 2B panels a vs. b and Figure 2C). We obtained and sequenced the construct that has been used in Dr Wigdahl's laboratory and the sequence results showed that the Tax1 NES domain had been cloned to the C-terminus part of the GFP rather than to the N-terminus (data not shown). Consequently, a second series of recombinant plasmids was made using the pGFP-C3 vector, allowing for a GFP C-terminal fusion construct. As with the NES2-EGFP construct, GFP-NES2 was mostly cytoplasmic (Figure 2B panel c), while, under these experimental conditions, GFP-NES1 was also, as previously published, preponderant in the cytoplasm (Figure 2B panel d). Interestingly, subcellular fractionation experiments clearly demonstrated that, even in that case, the GFP-NES1 nuclear fraction was more abundant than that of GFP-NES2 (Figure 2C right panel). Altogether, these results suggest that, without the context of a full length protein, Tax2 NES domain is active both when fused to the N- or to the C-terminus part of the GFP, while Tax1 NES functions more efficiently when fused to the C-end of GFP.
Within Tax2 NES sequence, several leucine residues are critical for a CRM-1 dependent function
We next investigated whether Tax2 NES activity was dependent upon the CRM-1 pathway. To this end, Hela cells were transfected with the different NES constructs (Figure 3A), with or without leptomycin B (LMB). LMB blocks CRM1-dependent nuclear export and has been used extensively to probe this process [36]. In the presence of LMB, GFP-NES2 localizes to the nucleus, suggesting that the CRM-1 pathway is involved in the shuttling of the fusion protein (Figure 3B panel b). As a control, incubation of the GFP-NES transfected cells with methanol (the solvent which has been used to dissolve LMB in panel b), had no effect (Figure 3B panel a). Leucine 195 (formerly named 194 [27]) has been shown to be critical for the NES1 ability to export the GFP protein via the CRM-1 dependant pathway. Since the sequence of NES2 also contains a leucine at position 195, we mutated this residue to an alanine. This mutation abrogated the nuclear export of GFP-NES2 (Figure 3B, panel c). As expected, adding LMB to the GFP-NES2 L195A had no effect on the protein localization (Figure 3B, panel d). Mutating leucine 200 to an alanine also suppressed NES function (Figure 3D), while the L194A mutation had no effect (data not shown). Altogether these results confirm that, within the Tax2 NES domain, more than one leucine residues are needed for the function of the export signal. Western blot controls show that the overall protein expression was comparable between the different constructs (Figures 3C and 3E).
Figure 3 Nucleocytoplasmic distribution of GFP-NES2 is altered after incubation with leptomycin B or by single point mutations. (A): Sequence alignment of wild-type Tax2 and Tax2 GFP-NES mutants. (B and D): HeLa cells were transiently transfected with the different GFP-NES plasmids. Eighteen hours post transfection, transfected cells were treated with leptomycin B (40 nM) or methanol for 3 hours. Cells were then washed, fixed, mounted with DAPI-containing medium and visualized with a Zeiss Axioplan 2 imaging microscope X40 using a Zeiss Axiocam HRc (color) camera and the Zeiss Apotome software. Images of cells that are representative of the entire population are shown. (C and E): Western-blot analysis of GFP and GFP-NES proteins. 293T cell lysates (70 μg) were subjected to electrophoresis on a 10 % TG gel and probed with GFP or β-tubulin antibodies.
Evaluating the role of Tax2 leucine 188
The NES prediction software results suggested that L188 might be part of NES2 (Figure 1). To evaluate the role of this amino acid in Tax2 NES function we constructed another series of NES-EGFP plasmids, in which amino acid leucine #188 was added to the autologous (i.e. NES2) or to the heterologous NES (i.e. NES1) sequences. The constructs were transfected into Hela cells and the expression of the fusion proteins determined by western-blot (Figure 4A). As described above, the NES1-EGFP has a stronger nuclear localization than NES2-EGFP (Figure 4B panels a vs. c). This is correlated with the fractionation experiment (Figure 4C left panel). Remarkably, adding a leucine to the NES1-EGFP sequence improved the "NES" phenotype, since the nuclear fraction is lower in the presence of leucine 188 (Figure 4B panel a vs. b and Figure 4C right panel). Adding leucine 188 to the 189NES2202-EGFP sequence increased only modestly the percentage of cells in which the signal was cytoplasmic (Figure 4B, panels c vs. d). Altogether, these results demonstrate that a leucine residue at position 188 allows a better export of the NES1-EGFP fusion protein. However, this leucine is dispensable in the context of a GFP-NES1 protein.
Figure 4 The presence of leucine 188 restores NES1 function within the context of a NES1-EGFP protein. (A): Western-blot analysis of the different EGFP and NES-EGFP proteins. 293T cell lysates (70 μg) were subjected to electrophoresis on a 10 % TG gel and probed with GFP or β-tubulin antibodies. (B): HeLa cells were transiently transfected with the different NES-EGFP plasmids. Eighteen hours post transfection, transfected cells were washed, fixed, mounted with DAPI-containing medium and visualized with a Zeiss Axioplan 2 imaging microscope X40 using a Zeiss Axiocam HRc (color) camera and the Zeiss Apotome software. Images of cells that are representative of the entire population are shown. (C): 293T nuclear and cytoplasmic cell fractions were subjected to electrophoresis on a 10 % TG gel and probed with a GFP antibody. The western-blot results are representative of four independent experiments.
The localization of GFP-Tax2 is not altered by mutations in the NES
Although the results presented above have shown that the Tax2 NES represents an active domain in the context of the NES2-EGFP and GFP-NES2 chimera proteins, it was important to examine the function of the NES2 within the context of the full-length Tax2 protein. To do this, point mutations were made in the GFP-Tax2 full-length construct, with one or several leucine residues (up to three) mutated within the NES2. We also mutated residue 188 in order to evaluate the role of this amino-acid in the context of the complete protein. Unexpectedly, all these mutated Tax proteins, i.e. GFP-Tax2 L188Y, GFP-Tax2 L188Y, L191A, GFP-Tax2 L188Y, LL194–195AA, GFP-Tax2 L200A, had a predominant cytoplasmic localization and behaved mostly like Tax2 wild-type, i.e. with a strong cytoplasmic localization (Figure 5A, panels c, d, e, f as compared to b), but not like Tax1 (Figure 5A, panel a). Western-blot analysis demonstrated a comparable level of protein expression (Figure 5B). These results suggest that the presence of a wild-type NES2 is dispensable for exiting the cell nucleus in the context of the full-length Tax2 protein. As a control, we also used a GFP-Tax1 L200A vector. Strikingly however, in our hands this protein had a localization that was very similar to that of wild-type Tax-1 i.e. strong nuclear signal (data not shown). Indeed, we did not observe a strong localization to the nuclear membrane as it has been previously described. However, we should point out that we have used a GFP-Tax1 construct, while Alefantis et al used a Tax1-EGFP [27]. We cannot rule out the fact that the positioning of Tax1 L200A vis-à-vis GFP plays a role in the protein localization, although this is unlikely, since we previously observed that the localization of GFP-Tax1 was similar to that of Tax1-GFP [16].
Figure 5 Within the context of a full length Tax2 protein, the presence of a functional NES2 domain is dispensable for the protein localization. (A): HeLa cells were transiently transfected with the GFP-Tax1, GFP-Tax2 and the different GFP-Tax2 mutants plasmids. Eighteen hours post transfection, the cells were washed, fixed, mounted with DAPI-containing medium and visualized with a Zeiss Axioplan 2 imaging microscope X40 using a Zeiss Axiocam HRc (color) camera and the Zeiss Apotome software. Images of cells that are representative of the entire population are shown. (B): Western-blot analysis of GFP and GFP-NES proteins. 293T cell lysates (70 μg) were subjected to electrophoresis on a 10 % TG gel and probed with GFP or β-tubulin antibodies.
Discussion
Both in infected and in transfected cells, Tax1 and Tax2 are found in the nucleus and in the cytoplasm in different proportions: Tax1 being more abundant in the nucleus, while Tax2 is more prone to be found in the cytoplasm [16,35]. In the nucleus, Tax1 and Tax2 interact with transcription factors and activate the cyclic-AMP response element and activating transcription factor (ATF) binding (CREB/ATF) pathway, while in the cytoplasm the viral transactivators interact with several members of the NF-κB transduction pathway [5,37]. Tax1/Tax2 activation of CREB/ATF is needed for an efficient viral gene expression, while the permanent activation of NF-κB has been suggested to be critical, at least in HTLV-1 infected cells, for evading apoptosis. In order to activate the CREB/ATF and NF-κB pathways, both Tax1 and Tax2 must therefore shuttle between these two compartments [34].
A Nuclear Export Signal (amino acid 189 to 202) has recently been described in Tax1 [27]. Amino acids 1 to 58 constitute non canonical Nuclear Localization Signals [16,32] in both Tax1 and Tax2, but amino acids 90 to 100 are also critical for the localization of the viral transactivators [16]. Using prediction software as well as in vitro assays, we now describe another domain of Tax2. This sequence represents a Nuclear Export Signal (NES), with different functional characteristics from that of NES1. For example, the percentage of the GFP-NES2 protein that is present in the nucleus of the transfected cells is slightly different from that of GFP-NES1 protein. In addition, Tax2 NES is functional, no matter if it is fused to the N-terminal or the C-terminal of GFP, which is not the case of NES1 which is more active when fused to the C-terminus of GFP. We have also determined here that the NES of Tax2 can direct nuclear export via the CRM1 pathway, and that point mutations at positions 195 and 200 abrogate NES mediated translocation. All in all, these results demonstrate that the NES sequences of Tax1 and Tax2 have different functional profiles reflecting their slightly different sequences, and that the divergent amino acids are likely to be critical for the NES activity. The predictor software suggested that, in the Tax2 sequence, leucine 188 might also be part of the NES domain. This leucine is absent from Tax1 and, strikingly, when added to the NES1-EGFP construct, it restores the function of the Tax1 NES.
However, the most important result of this study is that, within the context of the whole Tax2 protein, mutating one or several leucine residues has no or an extremely limited impact on Tax2 localization. This could have been indicative of a secondary NES in the sequence being able to mediate translocation on its own, but this theory is not supported by the NetNES computational analysis. Therefore, this hypothesis is very unlikely. It would also disagree with our report that LMB treatment of Tax2 transfected cells did not abolish protein translocation [16]. Hence, we consider that the very modest increase in the GFP-Tax2 nuclear signal observed with some GFP-Tax2 mutants constructs as compared to GFP-Tax2 is not consistent with a strong use of this NES sequence by Tax2. Altogether, these results imply that the Tax2 protein uses other means of export from the cell nucleus leading to the observed strong cytoplasmic signal. This is consistent with our previous results showing that the 90–100 Tax domain, which does not behave as a NES, is critical for the protein localization [16]. Our results are therefore paradoxical: while Tax1 possesses a NES domain, it localizes predominantly in the nucleus at the equilibrium, whereas Tax-2, whose NES sequence is dispensable, has a predominant cytoplasmic localization.
In conclusion, without the context of the protein, both Tax1 and Tax2 seem to possess working nuclear export signals. If one regards the nuclear localization signal and nuclear export signal as competing forces, the Tax2 NES seems to be a more efficient mediator than that of Tax1 in terms of cytoplasmic versus nuclear abundance of the proteins without the context of a full-length protein. This observation is supported by the computational analysis as well as by our in vitro data. However, Tax2 does not need its NES signal to relocate to the cytoplasm. Rather, it seems to employ a different, hitherto uncharacterized translocation system, as we have previously suggested [16]. The implications of this paradox are that, even though a fully functional nuclear export signal is embedded in the Tax2 sequence, it is not actually necessary for the protein translocation under the conditions tested here. However, its functional conservation suggests that it might have a biological impact on the protein functions. Future in vivo studies will decipher whether the presence of the "NES" sequence in the HTLV-2 Tax protein has any role during the viral cycle.
Methods
Cell culture and drug treatment
Hela and 293T cells were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum and antibiotics (penicillin 100 U/ml and streptomycin at 100 μg/ml). Cell lines were maintained at 37°C in 5% CO2. When indicated, cells were incubated with leptomycin B (Sigma) at 40 nM for 3 h.
GFP-NES, NES-EGFP and GFP-Tax protein construction
The GFP-NES and NES-EGFP recombinants plasmids were obtained by cloning double stranded oligonucleotides into GFP-C3 and EGFP-N1 vectors (Clontech), using SacI/EcoRI and XhoI/PstI restriction sites respectively. Single or combined point mutations (at amino acids 188, 191, 194, 195 and 200) were also made in GFP-Tax1 and GFP-Tax2 sequences using the quick change mutagenesis kit (Stratagene) [16]. The nucleotide sequences of all constructs were determined using the DYEnamic ET Terminator Cycle Sequencing Kit (Amersham Biosciences) on an Applied Biosystems 373A DNA sequencer. Of note, during the course of these experiments, we noticed that the amino-acid numbering that has been used in Alefantis article was incorrect [27]. The first lysine of the Tax1 NES sequence is at position 189 and not 188 as reported previously. We have therefore modified the amino acid number accordingly.
Transient transfection
For microscopic analyses, Hela cells were seeded in eight-well chamber glass slides, at a concentration of 3 × 104 cells/well and transfected the next day with 0,3 μg of DNA using the Effectene reagent (Qiagen). For immunoblot analyses, 293T cells were seeded on 6-well plates at 6 × 105 cells/well and transfected the next day with 2 μg of DNA using the Polyfect reagent (Qiagen) following the manufacturer's instructions.
Immunoblot analyses
Twenty-four hours after transfection, 293T cells were washed twice with PBS, lysed (Tris-HCl pH 7,4 50 mM, NaCl 120 mM, EDTA 5 mM, NP40 0,5%, Na3VO4 0,2 mM, DTT 1 mM, PMSF 1 mM) in the presence of protease inhibitors (Complete, Boehringer) and incubated on ice. Cell debris were pelleted by centrifugation. Protein concentration was determined by Bradford (Biorad). Samples were loaded into 10% Tris/Glycine gels (Invitrogen) subjected to electrophoresis at 130V and transferred onto a nitrocellulose membrane (Immobilon-P, Millipore). Membranes were blocked in a 5% PBS-milk solution, incubated with a specific anti-GFP antibody (JL-8, BD 1:1000) overnight at 4°C. The next day, the membranes were washed and incubated with an anti-mouse horseradish peroxidase-conjugated secondary antibody (Amersham Biosciences 1:40000) and developed using the SuperSignal West Pico Kit (Pierce). To control for the amount of protein loaded per well, membranes were stripped with the Re-blot Plus Kit (Chemicon International), and re-probed with a specific anti β-tubulin antibody (sc9104 Santa Cruz Biotechnology 1:1000).
Green fluorescent protein analyses
Twenty-four hours after transfection, the cells were washed with PBS, fixed with 4% paraformaldehyde (Sigma) and washed with PBS. Nucleic acids were stained with 4'-6'-diamine-2 phenylindole dihydrochloride (DAPI)-containing mounting medium (Vectashield, Vector). Cells were visualized with a Zeiss Axioplan 2 imaging microscope X40 using a Zeiss Axiocam HRc (color) camera and the Zeiss Apotome software. Given the fact that the localization of the GFP-fusion proteins is similar in Hela and in 293T, and because 293T cells are complex to handle in immunofluorescence experiments, we used these cells only for the western-blot analyses.
Nuclear and cytoplasmic extraction
Twenty-four hours after transfection, the cells were washed with PBS. Nuclear and cytoplasmic fractions were then isolated using the sub-cellular proteome extraction kit (Calbiochem) following the manufacturer's instructions. Samples were subjected to immunoblot analyses as described above.
NetNES software analysis
This software predicts leucine-rich nuclear export signals (NES) in eukaryotic proteins using a combination of neural networks (NN) and hidden Markov models (HMM). The prediction server calculates a combined 'NES score' from the NN and HMM scores. If the calculated 'NES score' exceeds the threshold, then that particular residue is expected to participate in a nuclear export signal. This is denoted with a "Yes" in the column "Predicted". Of note, the reason why one gets different scores for the same residues when comparing Tax1 and Tax2 sequences is that the score depends not only on the residue in question, but also on a number of previous residues which, in the case of E193 for example, are not identical between the two sequences.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
SAC, SC and LM performed the laboratory work. AG was involved in drafting the manuscript. LK participated in the interpretation of the NetNES server results and helped drafting the manuscript. RM designed, implemented and coordinated the study and wrote the manuscript. All authors have read and approved the manuscript.
Acknowledgements
This work was funded by Institut Pasteur, by grants from l'Association de Recherche sur le Cancer (ARC # 4781) and from ARECA to RM, fellowships from le Ministère de la Recherche to SAC, from CANAM and Pasteur Weizmann fellowships to LM and from Association Virus Cancer Prévention and La Ligue Contre le Cancer to SC. RM is supported by INSERM.
==== Refs
Feuer G Green PL Comparative biology of human T-cell lymphotropic virus type 1 (HTLV-1) and HTLV-2 Oncogene 2005 24 5996 6004 16155606 10.1038/sj.onc.1208971
Gessain A Barin F Vernant JC Gout O Maurs L Calender A de The G Antibodies to human T-lymphotropic virus type-I in patients with tropical spastic paraparesis Lancet 1985 2 407 410 2863442 10.1016/S0140-6736(85)92734-5
Murphy EL Fridey J Smith JW Engstrom J Sacher RA Miller K Gibble J Stevens J Thomson R Hansma D Kaplan J Khabbaz R Nemo G HTLV-associated myelopathy in a cohort of HTLV-I and HTLV-II-infected blood donors. The REDS investigators Neurology 1997 48 315 320 9040713
Poiesz BJ Ruscetti FW Gazdar AF Bunn PA Minna JD Gallo RC Detection and isolation of type C retrovirus particles from fresh and cultured lymphocytes of a patient with cutaneous T-cell lymphoma Proc Natl Acad Sci U S A 1980 77 7415 7419 6261256
Grassmann R Aboud M Jeang KT Molecular mechanisms of cellular transformation by HTLV-1 Tax Oncogene 2005 24 5976 5985 16155604 10.1038/sj.onc.1208978
Jeang KT Giam CZ Majone F Aboud M Life, death, and tax: role of HTLV-I oncoprotein in genetic instability and cellular transformation J Biol Chem 2004 279 31991 31994 15090550 10.1074/jbc.R400009200
Mahieux R Pise-Masison CA Lambert PF Nicot C De Marchis L Gessain A Green P Hall W Brady JN Differences in the ability of human T-cell lymphotropic virus type 1 (HTLV-1) and HTLV-2 tax to inhibit p53 function J Virol 2000 74 6866 6874 10888626 10.1128/JVI.74.15.6866-6874.2000
Mahieux R Pise-Masison CA Nicot C Green P Hall WW Brady JN Inactivation of p53 by HTLV type 1 and HTLV type 2 Tax trans-activators AIDS Res Hum Retroviruses 2000 16 1677 1681 11080809 10.1089/08892220050193137
Ross TM Minella AC Fang ZY Pettiford SM Green PL Mutational analysis of human T-cell leukemia virus type 2 Tax J Virol 1997 71 8912 8917 9343258
Ross TM Pettiford SM Green PL The tax gene of human T-cell leukemia virus type 2 is essential for transformation of human T lymphocytes J Virol 1996 70 5194 5202 8764028
Semmes OJ Majone F Cantemir C Turchetto L Hjelle B Jeang KT HTLV-I and HTLV-II Tax: differences in induction of micronuclei in cells and transcriptional activation of viral LTRs Virology 1996 217 373 379 8599225 10.1006/viro.1996.0126
Sieburg M Tripp A Ma JW Feuer G Human T-cell leukemia virus type 1 (HTLV-1) and HTLV-2 tax oncoproteins modulate cell cycle progression and apoptosis J Virol 2004 78 10399 10409 15367606 10.1128/JVI.78.19.10399-10409.2004
Tanaka Y Hayashi M Takagi S Yoshie O Differential transactivation of the intercellular adhesion molecule 1 gene promoter by Tax1 and Tax2 of human T-cell leukemia viruses J Virol 1996 70 8508 8517 8970974
Tripp A Liu Y Sieburg M Montalbano J Wrzesinski S Feuer G Human T-cell leukemia virus type 1 tax oncoprotein suppression of multilineage hematopoiesis of CD34+ cells in vitro J Virol 2003 77 12152 12164 14581552 10.1128/JVI.77.22.12152-12164.2003
Tsubata C Higuchi M Takahashi M Oie M Tanaka Y Gejyo F Fujii M PDZ domain-binding motif of human T-cell leukemia virus type 1 Tax oncoprotein is essential for the interleukin 2 independent growth induction of a T-cell line Retrovirology 2005 2 46 16042787 10.1186/1742-4690-2-46
Meertens L Chevalier S Weil R Gessain A Mahieux R A 10-amino acid domain within human T-cell leukemia virus type 1 and type 2 tax protein sequences is responsible for their divergent subcellular distribution J Biol Chem 2004 279 43307 43320 15269214 10.1074/jbc.M400497200
Meertens L Pise-Masison C Quere N Brady J Gessain A Mahieux R Utilization of the CBP but not the p300 co-activator by human T-lymphotropic virus type-2 Tax for p53 inhibition Oncogene 2004 23 5447 5458 15156194 10.1038/sj.onc.1207719
Niinuma A Higuchi M Takahashi M Oie M Tanaka Y Gejyo F Tanaka N Sugamura K Xie L Green PL Fujii M Aberrant activation of the interleukin-2 autocrine loop through the nuclear factor of activated T cells by nonleukemogenic human T-cell leukemia virus type 2 but not by leukemogenic type 1 virus J Virol 2005 79 11925 11934 16140768 10.1128/JVI.79.18.11925-11934.2005
Endo K Hirata A Iwai K Sakurai M Fukushi M Oie M Higuchi M Hall WW Gejyo F Fujii M Human T-cell leukemia virus type 2 (HTLV-2) Tax protein transforms a rat fibroblast cell line but less efficiently than HTLV-1 Tax J Virol 2002 76 2648 2653 11861831 10.1128/JVI.76.6.2648-2653.2002
Fahrenkrog B Koser J Aebi U The nuclear pore complex: a jack of all trades? Trends Biochem Sci 2004 29 175 182 15082311 10.1016/j.tibs.2004.02.006
Xu L Massague J Nucleocytoplasmic shuttling of signal transducers Nat Rev Mol Cell Biol 2004 5 209 219 14991001 10.1038/nrm1331
Quensel C Friedrich B Sommer T Hartmann E Kohler M In vivo analysis of importin alpha proteins reveals cellular proliferation inhibition and substrate specificity Mol Cell Biol 2004 24 10246 10255 15542834 10.1128/MCB.24.23.10246-10255.2004
la Cour T Kiemer L Molgaard A Gupta R Skriver K Brunak S Analysis and prediction of leucine-rich nuclear export signals Protein Eng Des Sel 2004 17 527 536 15314210 10.1093/protein/gzh062
la Cour T Gupta R Rapacki K Skriver K Poulsen FM Brunak S NESbase version 1.0: a database of nuclear export signals Nucleic Acids Res 2003 31 393 396 12520031 10.1093/nar/gkg101
Daelemans D Costes SV Lockett S Pavlakis GN Kinetic and molecular analysis of nuclear export factor CRM1 association with its cargo in vivo Mol Cell Biol 2005 25 728 739 15632073 10.1128/MCB.25.2.728-739.2005
Yashiroda Y Yoshida M Nucleo-cytoplasmic transport of proteins as a target for therapeutic drugs Curr Med Chem 2003 10 741 748 12678777 10.2174/0929867033457791
Alefantis T Barmak K Harhaj EW Grant C Wigdahl B Characterization of a nuclear export signal within the human T cell leukemia virus type I transactivator protein Tax J Biol Chem 2003 278 21814 21822 12670929 10.1074/jbc.M211576200
Felber BK Paskalis H Kleinman-Ewing C Wong-Staal F Pavlakis GN The pX protein of HTLV-I is a transcriptional activator of its long terminal repeats Science 1985 229 675 679 2992082
Goh WC Sodroski J Rosen C Essex M Haseltine WA Subcellular localization of the product of the long open reading frame of human T-cell leukemia virus type I Science 1985 227 1227 1228 2983419
Kiyokawa T Kawaguchi T Seiki M Yoshida M Association of the pX gene product of human T-cell leukemia virus type-I with nucleus Virology 1985 147 462 465 3000074 10.1016/0042-6822(85)90149-7
Slamon DJ Press MF Souza LM Murdock DC Cline MJ Golde DW Gasson JC Chen IS Studies of the putative transforming protein of the type I human T-cell leukemia virus Science 1985 228 1427 1430 2990027
Smith MR Greene WC Characterization of a novel nuclear localization signal in the HTLV-I tax transactivator protein Virology 1992 187 316 320 1736534 10.1016/0042-6822(92)90320-O
Semmes OJ Jeang KT Localization of human T-cell leukemia virus type 1 tax to subnuclear compartments that overlap with interchromatin speckles J Virol 1996 70 6347 6357 8709263
Burton M Upadhyaya CD Maier B Hope TJ Semmes OJ Human T-cell leukemia virus type 1 Tax shuttles between functionally discrete subcellular targets J Virol 2000 74 2351 2364 10666266 10.1128/JVI.74.5.2351-2364.2000
Turci M Romanelli MG Lorenzi P Righi P Bertazzoni U Localization of HTLV-2 Tax protein is dependent upon a nuclear localization determinant in the N-terminal region Gene
Fornerod M Ohno M Yoshida M Mattaj IW CRM1 is an export receptor for leucine-rich nuclear export signals Cell 1997 90 1051 1060 9323133 10.1016/S0092-8674(00)80371-2
Sun SC Yamaoka S Activation of NF-kappaB by HTLV-I and implications for cell transformation Oncogene 2005 24 5952 5964 16155602 10.1038/sj.onc.1208969
|
16285885
|
PMC1308865
|
CC BY
|
2021-01-04 16:36:39
|
no
|
Retrovirology. 2005 Nov 14; 2:70
|
utf-8
|
Retrovirology
| 2,005 |
10.1186/1742-4690-2-70
|
oa_comm
|
==== Front
RetrovirologyRetrovirology1742-4690BioMed Central London 1742-4690-2-711629319310.1186/1742-4690-2-71ResearchInhibition of HIV derived lentiviral production by TAR RNA binding domain of TAT protein Mi Michael Y [email protected] Jiying [email protected] Yukai [email protected] Departments of Dermatology and Immunology, University of Pittsburgh, School of Medicine. 190 Lothrop St, Suite 145, Pittsburgh, PA 15261, USA2005 17 11 2005 2 71 71 31 7 2005 17 11 2005 Copyright © 2005 Mi et al; licensee BioMed Central Ltd.2005Mi et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
A critical step in the production of new HIV virions involves the TAT protein binding to the TAR element. The TAT protein contains in close proximity its TAR RNA binding domain and protein transduction domain (PTD). The PTD domain of TAT has been identified as being instrumental in the protein's ability to cross mammalian cell and nuclear membranes. All together, this information led us to form the hypothesis that a protein containing the TAR RNA binding domain could compete with the native full length TAT protein and effectively block the TAR RNA binding site in transduced HIV infected cells.
Results
We synthesized a short peptide named Tat-P, which contained the TAR RNA binding and PTD domains to examine whether the peptide has the potential of inhibiting TAT dependent HIV replication. We investigated the inhibiting effects of Tat-P in vitro using a HIV derived lentiviral vector model. We found that the TAT PTD domain not only efficiently transduced test cells, but also effectively inhibited the production of lentiviral particles in a TAT dependent manner. These results were also supported by data derived from the TAT activated LTR-luciferase expression model and RNA binding assays.
Conclusion
Tat-P may become part of a category of anti-HIV drugs that competes with full length TAT proteins to inhibit HIV replication. In addition, this study indicates that the HIV derived lentiviral vector system is a safe and reliable screening method for anti-HIV drugs, especially for those targeting the interaction of TAT and TAR RNAs.
==== Body
Background
The HIV TAT protein is a key regulator of viral replication [1]. Binding of the TAT protein to the TAR element, a 59 nt sequence at the 5' end of nascent RNA, is the first critical step for producing full length HIV RNA. The transcription of HIV RNA from both integrated and non-integrated HIV genome is dependent on TAT protein [2]. Thus, interruption of this TAT-TAR interaction has been considered as a possible way to inhibit HIV replication [3]. TAR RNA decoys have been shown to be able to interfere with the binding of TAT proteins to native TAR elements, thus inhibiting HIV replication [4-6]. However, delivery of oligonucleotides in vivo is not trivial. Conversely, small synthetic substances, or short TAT peptides mimicking the TAT and TAR RNA binding domains have been shown to be promising inhibitors of HIV replication [7,8]. Furthermore, a different fragment of the TAT protein could compete for the binding site of the CXCR4 receptor on T cells and inhibit HIV entry [9]. Recently, several research groups have identified the TAR RNA binding domain of the TAT protein to be an arginine rich region (aa 49–59) [10,11]. In addition, TAT has been found to contain a protein transduction domain (PTD) that is able to cross cell membranes to freely enter cells [12]. Furthermore, this TAT PTD also has the ability to deliver big and small molecules into target cells and cell nuclei [13-15]. We have found that the TAT PTD and the TAR RNA binding domain are located in the same region of the TAT protein. The close proximity of these two properties led us to hypothesize that the sequence of this region could serve as a decoy by competing with full-length native TAT proteins. Blocking the interaction between native TAT proteins and the TAR RNA could subsequently inhibit viral replication.
The lack of access to hazardous HIV laboratories has hindered anti-HIV drug development. For this reason, it is important to explore substitute HIV models. One option is to use non-human lentiviral models, such as equine infectious anemia virus (EIAV) [16], feline immunodeficiency virus (FIV) [17], bovine immunodeficiency virus (BIV) [18], and simian immunodeficiency virus (SIV) [19,20]. While these animal models have revealed important lentivirus replication and pathogenesis mechanisms, some discrepancies still exist between animal and human lentiviruses (HIV). For instance, the above animal models may not reflect the actual HIV life cycle in humans.
A different research method is represented by the HIV derived recombinant lentiviral vector system, which was developed for human gene therapy purposes [21]. First generation HIV based lentiviral vectors were generated by deleting the viral envelope gene (env) and replacing it with the vesicular stomatitis virus glycoprotein (VSV-G) gene to eliminate viral tropism for T lymphocytes and macrophages. In addition, gag, pol, and other regulatory HIV proteins were encoded on separate plasmids that were then co-transfected into the target cells. To improve on safety in second generation viral vectors, the accessory proteins encoding the nef, vif, vpu, and vpr genes were further deleted to reduce chances of generating replication competent recombinants [22]. However, the TAT and REV proteins were still required for producing lentiviral vectors and were provided by separate plasmids. In third generation lentiviral vectors, the introduction of strong chimeric promoters drove the full length RNA without the assistance of TAT [23]. Because second generation lentiviral vectors are dependent on TAT, we should be able to design experiments to examine anti-HIV approaches that target the TAT protein. Simultaneously, the third generation lentiviral vectors that are TAT independent can be used as controls. As described above, the use of theses vectors represent a strong biosafety profile. Additionally, by coding a marker gene into the recombinant lentiviral vector model, such as green fluorescent protein (GFP), we can easily measure viral infectivity and titer through cell counts, rather than measuring viral load indirectly through p24 or other viral structural products.
In this study, we describe the synthesis of a short peptide named Tat-P, which shares the same sequence as the TAR-RNA binding domain and the TAT PTD domain, and this peptide was evaluated in vitro using the HIV derived recombinant lentiviral vector model to examine its potential for inhibiting TAT dependent HIV replication. The ultimate goal of these studies was to determine if Tat-P could cross cellular and nuclear membranes and effectively block native TAT proteins from binding to TAR-RNA.
Results
Tat-P and Con-P1 peptides efficiently transduced 293T cells
In order to prevent native TAT proteins from binding to TAR-RNA, Tat-P must have the capability of crossing cell and nuclear membranes. To assess the transduction efficiency of Tat-P and two control peptides, Con-P1 and Con-P2, we synthesized FITC conjugated peptides. Con-P1 was utilized as a positive control because previous studies have demonstrated that this peptide shares similar structure and cell entry properties to Tat-P, conversely, Con-P2 represented a negative control because it lacks the PTD domain and its associative cell entry capabilities [24]. The 293T cells were treated with FITC labeled peptides ranged from 6.25 μM to 200 μM for 3 hours at 37°C, and internalization of these peptides was evaluated by fluorescent microscopy. As shown in Fig. 1A, the 293T cells displayed high levels of transduction by both Tat-P and Con-P1, and that the degree of transduction for these peptides was observed to be dose dependent. Furthermore, the peptide was found in the nucleus of transduced cells when examined with confocal microscopy (Fig. 1B), suggesting that the peptide was indeed inside the cells not simply attached to the cell surface. As expected, the Con-P2 negative control peptide was unable to transduce the 293T cells. These data confirm previous reports that Tat-P can cross cell membranes to enter the cytoplasm and then the nucleus.
Figure 1 Transduction of 293T cells by Tat-P and Con-P1. To test the capability of Tat-P to cross 293T cell membranes, FITC labeled Tat-P, Con-P1 or Con-P2 peptides were added to 293T cells with concentrations ranged from 6.25 μM and 200 μM. Three hours later, cells were washed extensively with PBS and viewed under fluorescent microscope (Magnification ×200) (Panel A). In some experiments, after transduction with 200 μM peptides, 293T cells were fixed and the nuclei were counter-stained with Sytox Orange (red). Cells were then visualized under confocal microscope (Magnification ×1000) (Panel B).
Tat-P inhibited the viral production of second-generation recombinant lentiviral vector
To evaluate the blocking of HIV TAT and TAR RNA interaction as a feasible target for anti-HIV drug development and to test whether the Tat-P blocks lentiviral vector particles production, 293T cells were transfected with three plasmids providing necessary genes to package replication-defective pseudo-typed HIV particles. Twenty-four hours after the transfection, Tat-P or the control peptides Con-P1 and Con-P2, were added to the cells. If Tat-P is able to compete with full length TAT protein for binding to TAR-RNA, it should block TAT transactivation activity and thus inhibit the viral RNA transcription and lentiviral production. We utilized the following two indicators to evaluate the inhibition of recombinant lentiviral production.
(1) Visualization of HIV production by electronic microscopy (EM)
Twenty-four hours following transfection, the media was replaced with fresh media containing 200 μM of Tat-P, Con-P1, or the same amount of PBS for 12 hours. The cells were then fixed and sectioned for transmission EM imaging. Fig. 2A shows that HIV particles were formed by the Tat-P and Con-P1 transfected 293T cells. From these EM images, the recombinant lentiviral vectors were visualized as 80~100 nm enveloped viral particles. It is important to note that the Tat-P treated cells showed formation of fewer viral particles than those of Con-P1 and the PBS treated controls.
Figure 2 Inhibition of recombinant lentiviral vector generation by Tat-P peptide. Panel A. To visualize the genetically generated pseudo HIV particles, 293T cells were co-transfected for 24 hours with 3 plasmids: pCMV Δ8.91, pMD VSV-G, and pHR'GFP. Then 200 μM of Tat-P and Con-P1 peptides, and the same amount of PBS, were added to the 293T cells for 12 hours. The cells were then fixed and sectioned for EM imaging (Magnification × 60,000). Arrows indicate the virus particles. Panel B and Panel C. To examine the Tat-P inhibition of HIV derived recombinant lentiviral vectors, 293T cells were cotransfected with 3 plasmids for 24 hours. The medium was replaced with DMEM containing 200 μM of Tat-P, Con-P1, Con-P2 peptides or PBS. Six hours later, the supernatants containing the viral particles were collected and the vector titers were determined. A representative from three individual experiments is presented. Panel D. Percent inhibition was calculated using the formula (1-titer in the presence of peptide/titer in the presence of PBS) × 100.
(2). Reduction in lentiviral titers following addition of the Tat-P
To accurately assess the Tat-P inhibition capability, we measured the lentiviral vector titer in the cell culture supernatant generated from co-transfection in the presence or absence of peptides. As shown in Fig. 2B, cell culture supernatant from co-transfection in the presence of Tat-P generated significantly fewer number of GFP positive cells, indicating much lower lentiviral vector titer in the preparation. The vector titer was calculated based on the initial number of 293T cells when lentiviral vector was added. The lentiviral vector titer was dramatically reduced in the presence of Tat-P (Fig. 2C). The inhibition effects of Tat-P, Con-P1, and Con-P2 on the lentiviral production were calculated to be 89.8%, 4%, and 5.9% compared to PBS control, respectively (Fig. 2D), suggesting that Tat-P strongly inhibit the lentiviral production in the setting of three plasmids co-transfection approach.
Tat-P inhibition of virus production was constant over time and the degree of inhibition was dose dependent
Tat-P inhibition effect was also demonstrated in the time point of 12 hours and 24 hours after addition of the peptide. As shown in Fig. 3A, the inhibition rates were 83% and 79%, respectively. The inhibition effect was decreased with incremental time length possibly due to the peptide degradation. In contrast, Con-P1 peptide had no inhibition effect at all time points, further suggesting the inhibition by Tat-P was specific. To evaluate the dose-response effect of the Tat-P on viral production, three different doses of Tat-P (200 μM, 100 μM and 50 μM) were utilized in the experiment. At each dose of the treatment, Tat-P inhibited the viral production quantified by flow cytometry (Fig. 3B) when compared to Con-P1 treatment and PBS control (data not shown). Compared to Con-P1 peptide, the inhibition rates of Tat-P were calculated to be 87.1% at 200 μM, 72.7% at 100 μM, and 59.2% at 50 μM.
Figure 3 Time and dose effects of Tat-P on recombinant lentiviral vector production. Panel A. To investigate the Tat-P driven inhibition of lentiviral vector generation over time, supernatants from the viral particle producing 293T cell cultures, in the presence of 200 μM of Tat-P, Con-P1, Con-P2, and PBS, were collected at 6 hour, 12 hour, and 24 hour time points. The supernatants containing the viral particles were added to freshly cultured 293T cell and these supernatants were evaluated for virus titers. Panel B. To assess dose-dependency of the Tat-P inhibition activity, supernatants from the viral particle producing 293T cell cultures, in the presence of 200 μM, 100 μM, and 50 μM of Tat-P and Con-P1, were collected at the 6-hour time point. Viral titers were determined and the inhibition effects were calculated.
Tat-P did not inhibit third generation virus production
In this experiment, we evaluated whether the inhibition of HIV replication by Tat-P was TAT protein dependent. Since the TAT protein is not required to produce third generation recombinant lentiviral vectors, then Tat-P should not inhibit third generation viral production. 293T cells were co-transfected within a third generation (TAT independent) lentiviral vector system. After exposure to Tat-P, Con-P1, Con-P2 and PBS, the cell supernatants were measured to determine virus titers (Fig. 4). All three peptides showed low levels (<10%) of virus replication inhibition. These data strongly support that the Tat-P inhibition of virus replication present above was occurring through direct interference with the native TAT proteins and their target TAR-RNA.
Figure 4 No inhibition of third generation lentiviral vectors by Tat-P was observed. To test that the Tat-P inhibition activity is specifically targeting HIV TAT protein, 293T cells were cotransfected with four plasmids of a third generation lentiviral vector system that is independent of the TAT protein. Then, 200 μM of Tat-P, Con-P1, and Con-P2 peptides, or a PBS control were added to the 293T cells for 6 hours. The supernatants containing the viral particles were collected and added to freshly cultured 293T cells to measure viral titers.
Tat-P toxicity of 293T cells did not occur at concentrations less than 400 μM
To evaluate the cell toxicity of Tat-P, escalating doses of the peptides were applied to 293 T cells. The cell viability was measured using MTT assay. Fig. 5 showed that the addition of Tat-P to the cell culture medium did not affect 293T cell viability up to 200 μM. A low level of toxicity was observed when the peptide concentration reached 400 μM. However, this toxicity level was similar to that induced by control peptides Con-P1 and Con-P2, suggesting that the inhibition of recombinant lentiviral production Tat-P is not due to the effect of cell toxicity.
Figure 5 Cytotoxicity of Tat-P on 293T cells. To test for peptide toxicity, Tat-P, Con-P1, and Con-P2 peptides in concentrations ranging from 0 μM to 400 μM were added to 293T cells at 37°C for 6 hours, and the cell viabilities were monitored by MTT assay.
Tat-P Inhibits TAT Activated LTR-Luciferase Activity
To verify the results that Tat-P competitively inhibited HIV based lentiviral production via interference with TAR RNA binding, HIV LTR-luciferase expression model was established by cotransfection of 293T cells with pLTR-luc and pCMV-TAT plasmids. The expression of luciferase is augmented in the presence of full length TAT protein through the binding of TAR RNA. The binding of Tat-P to TAR RNA should competitively block the interaction of TAT protein with TAR RNA, resulting in reduction of reporter gene expression. As demonstrated in Fig. 6A, luciferase activity decreased in the presence of Tat-P in a dose dependent manner. In contrast, Con-P1 peptide has no effect of luciferase gene expression, suggesting that the inhibitory effect of Tat-P ensues from competition with TAT protein and does not represent nonspecific transcription inhibitory effect.
Figure 6 Inhibitory effect of Tat-P on TAT activated LTR-Luciferase activity and specific TAR RNA binding by Tat-P. Panel A. The TAT activated LTR-luciferase assay: 239T cells were co-transfected with the pLTR-luc and pCMV-TAT plasmids, and Tat-P and Con-P1 peptides (200 μM, 100 μM, 50 μM) were added to the cells 6 hours after transfection. The conditioned media were exchanged with fresh media containing the same amounts of peptides after 12 hours. The cells were harvested 6 hours later and processed by luciferase assay. The inhibition rates were expressed as mean ± SE. Panel B. The RNA binding assay: 0.25 nmol of TAR RNA was incubated with Tat-P or control peptides (Con-P1 and Con-P2) at indicated Peptide: TAR RNA molar ratio in a total 10 ul of reaction mixture for 15 minutes on ice. Free RNAs and peptide-RNA complexes were resolved by electrophoresis at 25°C on 15% polyacrylamide gels, and imaged using a fluorescent-based EMSA kit.
Tat-P Specifically Binds To TAR-RNA
We next investigated whether Tat-P's antiviral activity was due to specific binding to TAR-RNA by performing Tat-P and TAR-RNA binding assays in vitro. Tat-TAR-RNA complexes were formed by mixing serial dilutions of Tat-P peptide with TAR-RNA, and resolving the peptide-RNA complexes by electrophoresis on polyacrylamide gels. Fig. 6B shows that Tat-P peptides did bind to the TAR-RNA and these complexes are represented by upward shifts in the gel. As the concentration of Tat-P increased (left to right), the RNA bands showed a continuous step-up pattern indicating increasing density. No such phenomenon was observed for the control peptides, suggesting that the Tat-P peptides were binding specifically to the TAR-RNA.
Discussion
Currently, treatments for HIV infection rely heavily on anti-viral therapies. Most of these therapies target the HIV reverse transcriptase and protease enzymes by using nucleoside analogues as enzymes inhibitors, and their combination, known as highly active antiretroviral therapy (HAART), has markedly decreased mortality and morbidity in the developed world. The disadvantages of HAART include its inability to completely eradicate HIV from the body, long-term toxicity, and eventually the emergence of drug-resistant HIV strains [25]. Furthermore, the majority of HIV carriers have limited access to anti-retrovirals (ARVs) because of high costs and problems with patient compliance. It is, therefore, vital to find new strategies for identifying anti-HIV remedies, such as new targets of viral replication, new sources of drugs, and safe anti-HIV drug screening models.
Interruption of the formation of TAT-TAR-RNA complex represents such an endeavor. Small molecules mimicking either the native TAT peptides or TAR-RNA decoys have been investigated as new approaches for inhibiting HIV replication [4-9]. The lack of access to hazardous HIV laboratories is one of major hurdles for developing anti-HIV drugs. One option to overcome this restriction is to develop lower-risk assays for use in BSL-2 laboratories. Recombinant lentiviral vectors, widely used for gene therapy research could offer a potential substitute model for evaluating the efficacy of anti-HIV drugs. This may especially be true for candidate drugs targeting the interaction between TAT and TAR-RNA, the interaction of which is required for producing second generation recombinant lentiviral vectors. Based on the observation that the short Tat-P peptide can freely enter cells and specifically bind to TAR-RNA, we investigated the hypothesis that HIV replication could be inhibited by Tat-P peptides blocking native TAT proteins from binding to the TAR-RNA, and that these studies could be performed using HIV derived lentiviral model.
In these studies, we found that Tat-P was able to transduce 293T cell membranes without significant toxicity, and that the peptides inhibited recombinant lentiviral production in a TAT dependent manner. The inhibition of recombinant lentiviral production by Tat-P likely resulted from the competitive binding with TAR-RNA and the blocking of full length TAT by Tat-P. As demonstrated in Fig. 6B, Tat-P could bind to TAR RNA. More importantly, luciferase gene expression from TAT responsive LTR promoter was inhibited by the presence of Tat-P (Fig. 6A), further suggesting that the inhibition effect of Tat-P is mediated by interference with TAT-TAR RNA interaction. Compared to the dramatic inhibition of infective lentiviral particles (Fig. 3), the inhibition of luciferase gene expression from TAT responsive promoter by Tat-P seems less dramatic (Fig. 6A). Such discrepancy was also observed previously by others using TAT responsive promoter driven CAT assay (7). One possible explanation for the difference is that there is a higher amount of TAT protein may be produced from co-transfected plasmid pCMV-TAT. Thus, the same amount of Tat-P result in less effective competitive inhibition. In contrast, TAT protein level in the generation of viral particles by co-trasnfection method may be lower since it is generated by polycistronic mRNA from plasmid pCMV Δ8.91. Alternatively, viral particle production is a multiple steps process dependent on TAT. The competitive inhibition of TAT function by Tat-P may be amplified in the subsequent steps, resulting in more dramatic reduction of infective viral particles. In addition, it is possible that production of longer RNA is more dependent on the action of TAT, whereas the shorter luciferase gene expression from LTR promoter may be less dependent on TAT. Therefore, competitive blocking of TAT interaction with TAR RNA by Tat-P results in less dramatic inhibition of luciferase activity.
The recombinant lentiviral vector model has two advantages over natural HIV cell culture model. First, it is safer and able to be conducted in most laboratories. Second, it is an alternative approach for evaluating the infective recombinant viral particles. However, it is not clear if this recombinant lentiviral vector system can also be used to screen other anti-HIV drugs, such as those that target reverse transcriptase and proteinase. The split of one HIV genome into three different plasmids in generating a lentiviral vector may create an artificial setting for studying viral pathogenesis, which may affect the anti-HIV mechanisms. Thus, the results obtained through this recombinant lentiviral vector system need to be validated by conventional in vitro cell culture screening methods. Nevertheless, our research has shown that the recombinant lentiviral vector in vitro generation model may provide an easy and safer assay for primary screenings of ARV drugs before moving on to more involved methods requiring restricted P3 facilities.
Conclusion
Based on the above results, we draw the following conclusions: Tat-P inhibits HIV derived lentiviral production by blocking native TAT proteins from binding to TAR-RNA; genetically generated HIV models can be applied to screen anti-HIV drugs before using the high risk wild type HIV models; the results obtained from a recombinant lentiviral vector in vitro model need to be validated using wild type HIV cell culture methods and animal models.
Methods
Peptides and RNA
Tat-P (47YGRKKRRQRRR57) [10,12], Con-P1 (RRQRRTSKLMKR) [24] that shares similar structure and transduction efficiency as Tat-P, and Con-P2 (ARPLEHGSDKAT) [24] that lacks the capability of cell transduction, were synthesized (Peptide Synthesis Facility, University of Pittsburgh) using standard fmoc chemistry, then cleaved and deprotected by stirring in a 95% TFA, 2.5% triisopropylsilane, 2.5% H2O solution. The peptides were purified by reverse phase high performance liquid chromatography to >90% purity. Lyophilized peptides were reconstituted in PBS before use. To generate FITC labeled peptides, the fluorescein moiety (Fl) was attached via an aminohexanoic acid spacer by treating a resin-bound peptide (1.0 mmol) with FITC (1.0 mmol) and diisopropyl ethyl amine (5 mmol) in dimethylformamide (DMF; 10 ml) for 12 h [26]. Cleavage from the resin was achieved by using 95:5 trifluoroacetic acid (TFA)/triisopropylsilane. Removal of the solvent in vacuo gave a crude oil that was triturated with cold ether. The crude mixture thus obtained was centrifuged, the ether was removed by decantation, and the resulting orange solid was purified by RP-HPLC (H 2O/CH3CN in 0.1% TFA). The TAR RNA 29mer 5'-GCCAGAUCUGAGCCUGGGAGCUCUCUGGC-3' [10] was purchased from Dharmacon (Lafayette, CO) and the RNA was purified with PAGE gel and desalted by the manufacturer.
Transduction of 293T cells by peptides
FITC labeled Tat-P, Con-P1, and Con-P2 peptides were added to 293T cells at concentrations ranged from 6.25 μM to 200 μM and incubated at 37°C for 3 hours. The cells were washed extensively with PBS (pH.7.2) to remove excess peptides. Transduction of cells was visualized under a fluorescent microscope. To determine if the peptides were actually inside the cells, we conducted confocal microscopy study by co-staining the transduced cells with nucleus staining. 293T cells were transduced with 200 μM peptides. Three hours later, the treated cells were washed with tris buffered saline (TBS, pH 7.4) and fixed with 2% of paraformaldehyde containing 0.1% of Triton X-100 (Sigma, St. Louis, MO). The nuclei were stained with 1:2000 of Sytox Orange (Molecular Probes, Eugene, OR) and the peptide intracellular uptake was examined by confocal microscopy.
In vitro generation of lentiviral vectors
The production of second and third generation recombinant lentiviral vectors was performed as described previously using a three- or four- plasmids cotransfection procedure [22,27]. For generating third-generation lentiviral vectors, 80% confluent 293T cells were transfected with plasmid DNA pLenti-EGFP-TRIP together with packaging plasmids, pLP1, pLP2, and pVSV-G, (Invitrogen, San Diego, CA) using the calcium phosphate precipitation method according to manufacturer's description (Stratagene, San Diego, CA). To produce second-generation VSV pseudo-typed lentiviral vectors, plasmid pCMV Δ8.91 expressing the core proteins and enzymes of HIV, plasmid pMD VSV-G providing the envelope protein of VSV-G, and plasmid pHR'GFP expressing the green fluorescence protein (GFP) were utilized to transfect 293T cells using the same method as above. Handling of viral vectors was according to the guideline of BSL-2+ laboratories established by the Recombinant DNA Committee of University of Pittsburgh.
Assays for Tat-P inhibition of HIV lentiviral production
Twenty-four hours after the three plasmid transfection, media were replaced with fresh media containing different concentrations of the peptides. Cell supernatants containing viral particles were collected at 6 hour, 12 hour, and 24 hour time points to determine the viral titers by transducing 293T cells. Media collected at different time points were diluted two fold with fresh media containing 8 μg/ml of polybrene and then added to 293T cells. Two days later, cells were collected and the transduced EGFP+ cells were analyzed using flow cytometry (BD Bioscience, CA). Percentage of transduction was calculated. The quantitative data collected were expressed as mean ± SD, and the viral inhibition rates were calculated by the formula: Inhibition rate = (1 - Number of Tat-P Treated Green Cells/Number of Green Cells of a Control) × 100%.
Visualization of viral particles using electronic microscope
Twenty-four hours after transfection, Tat-P, Con-P1, or PBS was added to the 293T cells for 12 hours. The cells were washed with PBS twice and fixed using 2% glutaraldehyde. Viral particles were examined by electronic microscope (EM) imaging.
MTT assay for cell viability
The 293T cells were treated with medium containing peptide concentrations ranging from 0 μM to 400 μM for 6 hours at 37°C. MTT (Sigma Chemical Co, St. Louis, MO) was added to the wells at a concentration of 50 μg/ml at 37°C for 3 hours. Subsequently, the medium was aspirated, and the insoluble formazan crystals were dissolved in a solution of 10% SDS. Absorbance readings were taken at λ = 570 nm with background subtracted at λ = 630 nm [28].
TAT dependent LTR-luciferase assay
To investigate if TAT dependent LTR-luciferase expression can be inhibited by co-delivering Tat-P, 293T cells were cotransfected with HIV LTR driven luciferase cDNA plasmid (pLTR-luc) and CMV driven full length TAT cDNA plasmid (pCMV-TAT) using a calcium phosphate precipitation method. Both plasmids are kindly provided by Dr. P. Gupta of the University of Pittsburgh, School of Public Health. At 6 hours following transfection, Tat-P and Con-P1 peptides (200 μM, 100 μM, 50 μM) were added to the cotransfected 293T cells, and the conditioned media were exchanged with fresh media containing same amounts of peptides after 12 hours. The cells were harvested 6 hours later and processed by luciferase assay (Promega, Madison WI) and the level of luciferase activity was determined at 24 hours using an illuminometer (AutoLumat LB 953, EG&G berthold). The data collected were expressed as mean ± SE, and the luciferase inhibition rate was calculated by a formula: Inhibition rate = (1 - Luminescent Units of Tat-P Treated/Luminescent Units of a Control) × 100%.
RNA-binding assay
RNA Binding assays were performed according to a previous report [29]. Briefly, peptides and RNA were incubated together for 15 minutes on ice in 10 μl of a binding reaction mixture containing 10 mM hepes/KOH (pH 7.5), 100 mM KCl, 1 mM MgCl2, 0.5 mM EDTA, 1 mM dithiothreitol. To determine relative binding affinities, 0.25 nmol of TAR-RNA were titrated with serial dilutions of Tat-P, Con-P1 and Con-P2 (Peptide/RNA molar ratios are 0, 0.25, 0.5, 0.75, and 1). Free RNAs and peptide-RNA complexes were resolved by electrophoresis at 25°C in 15% polyacrylamide gels with 1xTBE (90 mM Tris/45 mM boric acid/1 mM EDTA) and imaged by fluorescent based Electrophoretic Mobility Shift Assay (EMSA) kit (Molecular Probes, Eugene, OR).
List of abbreviations
HIV: Human immunodeficiency virus
TAR: Trans-activating response region
TAT: Transactivating regulatory protein
PTD: Protein transduction domain
RNA: Ribonucleic acid
Tat-P: TAT peptide
293T: A human kidney epithelial cell line
Con-P1: Control peptide one
Con-P2: Control peptide two
EM: Electron microscopy
PBS: Phosphate buffered saline
TBS: Tris buffered saline
GFP: Green fluorescent protein
MTT: 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide
HAART: Highly active antiretroviral therapy
ARV: Anti-retroviral
FITC: Fluorescein isothiocyanate
VSV-G: Vesicular stomatitis virus glycoprotein
CMV: Cytomegalovirus
EMSA: Electrophoretic mobility shift assay EMSA
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MM designed and performed most of the experiments and wrote the manuscript. JZ provided crucial technical help for the experiments. YH supervised experimental design, experiment processes, data interpretation and writing of the manuscript.
Acknowledgements
The authors acknowledge Biologic Imaging Center of University of Pittsburgh for preparing the EM pictures.
==== Refs
Jeang KT Xiao H Rich EA Multifaceted activities of the HIV-1 transactivator of transcription, Tat J Biol Chem 1999 274 28837 28840 10506122 10.1074/jbc.274.41.28837
Wu Y HIV-1 gene expression: lessons from provirus and non-integrated DNA Retrovirology 2004 1 13 15219234 10.1186/1742-4690-1-13
Bannwarth S Gatignol A HIV-1 TAR RNA: the target of molecular interactions between the virus and its host Curr HIV Res 2005 3 61 71 15638724 10.2174/1570162052772924
Michienzi A Li S Zaia JA Rossi JJ A nucleolar TAR decoy inhibitor of HIV-1 replication Proc Natl Acad Sci USA 2002 99 14047 14052 12376617 10.1073/pnas.212229599
Garbesi A Hamy F Maffini M Albrecht G Klimkait T TAR-RNA binding by HIV-1 Tat protein is selectively inhibited by its L-enantiomer Nucleic Acids Res 1998 26 2886 2890 9611232 10.1093/nar/26.12.2886
Banerjea A Li MJ Remling L Rossi J Akkina R Lentiviral transduction of Tar Decoy and CCR5 ribozyme into CD34+ progenitor cells and derivation of HIV-1 resistant T cells and macrophages AIDS Res Ther 2004 1 2 15813986 10.1186/1742-6405-1-2
Choudhury I Wang J Rabson AB Stein S Pooyan S Leibowitz MJ Inhibition of HIV-1 replication by a Tat RNA-binding domain peptide analog J Acquir Immune Defic Syndr Hum Retrovirol 1998 17 104 111 9473010
Hamy F Felder ER Heizmann G Lazdins J Aboul-ela F Varani G Karn J Klimkait T An inhibitor of the Tat/TAR RNA interaction that effectively suppresses HIV-1 replication Proc Natl Acad Sci USA 1997 94 3548 3553 9108013 10.1073/pnas.94.8.3548
Lohr M Kibler KV Zachary I Jeang KT Selwood DL Small HIV-1-Tat peptides inhibit HIV replication in cultured T-cells Biochem Biophys Res Commun 2003 300 609 613 12504126 10.1016/S0006-291X(02)02903-0
Zhao H Li J Jiang L Inhibition of HIV-1 TAR RNA-Tat peptide complexation using poly(acrylic acid) Biochem Biophys Res Commun 2004 320 95 99 15207707 10.1016/j.bbrc.2004.05.148
Ruben S Perkins A Purcell R Joung K Sia R Burghoff R Haseltine WA Rosen CA Structural and functional characterization of human immunodeficiency virus tat protein J Virol 1989 63 1 8 2535718
Schwarze SR Ho A Vocero-Akbani A Dowdy SF In vivo protein transduction: delivery of a biologically active protein into the mouse Science 1999 285 1569 1572 10477521 10.1126/science.285.5433.1569
Ho A Schwarze SR Mermelstein SJ Waksman G Dowdy SF Synthetic protein transduction domains: enhanced transduction potential in vitro and in vivo Cancer Res 2001 61 474 477 11212234
Schwarze SR Hruska KA Dowdy SF Protein transduction: unrestricted delivery into all cells? Trends Cell Biol 2000 10 290 295 10856932 10.1016/S0962-8924(00)01771-2
Schwarze SR Dowdy SF In vivo protein transduction: intracellular delivery of biologically active proteins, compounds and DNA Trends Pharmacol Sci 2000 21 45 48 10664605 10.1016/S0165-6147(99)01429-7
Chen C Weisz OA Stolz DB Watkins SC Montelaro RC Differential effects of actin cytoskeleton dynamics on equine infectious anemia virus particle production J Virol 2004 78 882 891 14694119 10.1128/JVI.78.2.882-891.2004
Burkhard MJ Dean GA Transmission and immunopathogenesis of FIV in cats as a model for HIV Curr HIV Res 2003 1 15 29 15043209 10.2174/1570162033352101
Tok JB Bi L Huang S A comparative binding study of modified bovine immunodeficiency virus TAR RNA against its TAT peptide Bioorg Med Chem Lett 2004 14 6101 6105 15546738 10.1016/j.bmcl.2004.09.075
Zink MC Clements JE A novel simian immunodeficiency virus model that provides insight into mechanisms of human immunodeficiency virus central nervous system disease J Neurovirol 2002 8 42 48 12491150 10.1080/13550280290101076
Kumar A Narayan O Immunization for long-term protection against AIDS using the macaque model Virology 2001 285 1 5 11414799 10.1006/viro.2001.0942
Naldini L Blomer U Gallay P Ory D Mulligan R Gage FH Verma IM Trono D In vivo gene delivery and stable transduction of nondividing cells by a lentiviral vector Science 1996 272 263 267 8602510
Zufferey R Nagy D Mandel RJ Naldini L Trono D Multiply attenuated lentiviral vector achieves efficient gene delivery in vivo Nat Biotechnol 1997 15 871 875 9306402 10.1038/nbt0997-871
Dull T Zufferey R Kelly M Mandel RJ Nguyen M Trono D Naldini L A third-generation lentivirus vector with a conditional packaging system J Virol 1998 72 8463 8471 9765382
Mi Z Mai J Lu X Robbins PD Characterization of a class of cationic peptides able to facilitate efficient protein transduction in vitro and in vivo Mol Ther 2000 2 339 347 11020349 10.1006/mthe.2000.0137
Pereira CF Paridaen JT Anti-HIV drug development – an overview Curr Pharm Des 2004 10 4005 4037 15579085 10.2174/1381612043382459
Wender PA Mitchell DJ Pattabiraman K Pelkey ET Steinman L Rothbard JB The design, synthesis, and evaluation of molecules that enable or enhance cellular uptake: peptoid molecular transporters Proc Natl Acad Sci USA 2000 97 13003 13008 11087855 10.1073/pnas.97.24.13003
He Y Zhang J Mi Z Robbins P Falo LD Jr Immunization with lentiviral vector-transduced dendritic cells induces strong and long-lasting T cell responses and therapeutic immunity J Immunol 2005 174 3808 3817 15749922
Mai JC Mi Z Kim SH Ng B Robbins PD A proapoptotic peptide for the treatment of solid tumors Cancer Res 2001 61 7709 7712 11691780
Tan R Frankel AD Structural variety of arginine-rich RNA-binding peptides Proc Natl Acad Sci USA 1995 92 5282 5286 7777498
|
16293193
|
PMC1308866
|
CC BY
|
2021-01-04 16:36:41
|
no
|
Retrovirology. 2005 Nov 17; 2:71
|
utf-8
|
Retrovirology
| 2,005 |
10.1186/1742-4690-2-71
|
oa_comm
|
==== Front
Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-1331627448510.1186/1465-9921-6-133ResearchSequential analysis of surfactant, lung function and inflammation in cystic fibrosis patients Griese Matthias [email protected] Robert [email protected] Reinhold [email protected] Manfred [email protected] Karl [email protected] Ernst [email protected] Felix [email protected] Beat Study Group 1 Children's Hospital, University of Munich, Lindwurmstr 4, 80337 München, Germany2 Internal Medicine, University of Giessen, Klinikstr. 36, 35392 Giessen, Germany3 Department of Pediatric Pulmonology, Medical School, Carl-Neuberg-Str.1, 30625 Hannover, Germany4 Department of Pediatric Pulmonology and Immunology, Charité, Humboldt-University, Zum Heckeshorn 33, 14109 Berlin, Germany5 Department of Pediatric Pulmonology and Allergology, Children's Hospital, Josef Stelzmannstr.9, 50924 Köln, Germany6 Children's Hospital, University of Essen, Hufelandstrasse 55, 45122 Essen, Germany7 Principal investigators of the BEAT study group2005 7 11 2005 6 1 133 133 12 8 2005 7 11 2005 Copyright © 2005 Griese et al; licensee BioMed Central Ltd.2005Griese et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
In a cross-sectional analysis of cystic fibrosis (CF) patients with mild lung disease, reduced surfactant activity was correlated to increased neutrophilic airway inflammation, but not to lung function. So far, longitudinal measurements of surfactant function in CF patients are lacking and it remains unclear how these alterations relate to the progression of airway inflammation as well as decline in pulmonary function over time.
Methods
As part of the BEAT trial, a longitudinal study to assess the course of airway inflammation in CF, we studied lung function, surfactant function and endobronchial inflammation using bronchoalveolar lavage fluid from 20 CF patients with normal pulmonary function (median FEV1 94% of predicted) at three times over a three year period.
Results
There was a progressive loss of surfactant function, assessed as minimal surface tension. The decline in surfactant function was negatively correlated to an increase in neutrophilic inflammation and a decrease in lung function, assessed by FEV1, MEF75/25%VC, and MEF25%VC. The concentrations of the surfactant specific proteins A, C and D did not change, whereas SP-B increased during this time period.
Conclusion
Our findings suggest a link between loss of surfactant function driven by progressive airway inflammation and loss of small airway function in CF patients with limited lung disease.
Surfactant proteinphospholipidssurface activitypulsating bubblecapillary surfactometerbronchoalveolar lavageairway inflammationlung function
==== Body
Background
Cystic fibrosis (CF) is caused by mutations in the cystic fibrosis transmembrane regulator (CFTR) gene and early alterations are primarily found in the small airways [1]. Pulmonary surfactant is of critical importance for the maintenance of airspace patency during the respiratory cycle, especially at lower airway pressures at end-expiration. These biophysical functions are predominantly related to surfactant lipids [2,3], surfactant proteins (SP-) B and -C, as well as SP-A which contributes to resistance against inhibition of surfactant activity by products of inflammation. Altered surfactant has been associated with impaired lung function in young children with respiratory infections [4,5]. Thus, derangements of any of these components may affect the biophysical surfactant activity and may contribute to early alterations of lung function in patients with cystic fibrosis [6].
In a previous cross-sectional analysis of patients from the BEAT (Broncho-alveolar Lavage for the Evaluation of Anti-inflammatory Treatment) study, a long term study using bronchoalveolar lavage (BAL) to assess the evolution of airway inflammation in CF patients with normal lung function, we have observed a reduction of surfactant activity reflecting impaired airway patency in the majority of these CF patients [7]. The degree of surfactant dysfunction was inversely correlated with neutrophilic endobronchial inflammation, but no correlation was observed with flows at low lung volumes (MEF25%VC or MEF25/75%VC).
Flows at low lung volumes have been used to assess these early functional abnormalities, but due to inhomogeneous distribution of disease manifestation and a high inter-individual variability, they have not been found to be particularly sensitive when used cross-sectionally [8,9]. Longitudinal assessment of lung function is more sensitive in detecting deviations from personal best values [10,11]. Thus, sequential analyses appear to be more powerful to detect pathological changes [12,13]. We hypothesized that if surfactant function is relevant for small airway disease in CF patients, functional abnormalities in surfactant should be related to deterioration in pulmonary function assessed longitudinally. Some results of these studies have been reported previously in the form of an abstract [14].
Materials and methods
Patients and samples
Patients with cystic fibrosis were recruited from the BEAT study, a multi-center study using bronchoalveolar lavage to evaluate the evolution of airway inflammation in CF and its modulation by rhDNase treatment [15]. Inclusion criteria were an established diagnosis of CF, the ability to perform lung function tests, normal lung function defined as a FEV1 > 80% predicted, and absence of acute respiratory tract infections prior to bronchoscopy for at least 6 weeks. Patients receiving anti-inflammatory treatment (ibuprofen and systemic or inhaled corticosteroids), and those with allergic bronchopulmonary aspergillosis were excluded. The local ethic committees of all participating centers approved the study; written informed consent by both parents and/or patients was obtained in all cases. BAL was performed prior to randomization, after 1.5 as well as after 3 years in the same segment/subsegment over time. For this study samples from all patients were considered in whom 3 BALs were performed and sufficient material was available for analysis. More than 5 ml of BAL fluid at all 3 time points was available in 37 of 105 subjects. The measurement of surfactant proteins and functional assessment in the capillary surfactometer required 2.5 ml, therefore 2.5 ml could be used for surface tensions measurements in the bubble surfactometer. Since this technique requires at least 40 μg of phospholipids, phospholipid concentration had to be 16 μg/ml, which was the case in 22 subjects. Two samples of two different subjects were lost during processing; therefore 20 patients (13 females) with a median age of 10 years (range 5–16) were left for the final analysis. 13 of the patients were dF508 homozygeous, 3 were dF508 compound heterozygeous, in 2 patients other mutations were present, and in 2 the mutations were unknown. Four of the 20 patients were positive for P. aeruginosa. FVC was 91% pred. (70–126), FEV1 94% pred. (76–118), MEF25%VC 69% pred. (27–155)). The status of infection, i.e. the type of bacteria recovered from the airway specimens and serum values for C-reactive protein, total IgG and neutrophils were the same at all 3 time points assessed. The subpopulation did not differ from the total study population in both pulmonary function and baseline BAL parameters. 9 of 20 patients were randomized to rhDNase once daily. Because of the small number of patients in these subgroups, a sub-analysis related to rhDNase treatment was not performed in this study. For comparison to the levels in normal children, the mean and range of the biophysical and cellular measures in healthy children are indicated in the figures where appropriate. These and additional data obtained with the same methods can be found in a previous publication [7].
Bronchoalveolar lavage (BAL), sample preparation and biochemical and biophysical measurements
BAL was performed as described with 3 × 1 ml/kg body weight normal saline warmed to body temperature [15,16]. The first aliquot of the recovered BAL fluid was treated separately; all other samples were pooled for analysis. The total cell count was measured in a hemocytometer, the differential cell count was assessed from cytoprep slides. Aliquots of the cell free BAL supernatant of the pooled BAL sample were used for the analysis of total protein, and the surfactant proteins by ELISA (SP-A, SP-D [17] and SP-B, SP-C [18,19]). Pulmonary surfactant may be differentiated into a very surface active fraction, i.e. the large surfactant aggregates (LA) and a less active fraction containing smaller surfactant particles (small aggregates, SA), which may also inhibit surfactant function. 5 ml aliquots of the BAL fluid were centrifuged at 40,000 × g for 30 min to generate a surfactant pellet, containing the LA and the supernatant, containing the SA [19]. The phospholipid content was determined by phosphorus assay of a lipid extract of the surfactant pellet [19]. After resuspension of the surfactant pellet (NaCl 140 mM, HEPES 10 mM, EDTA 0.5 mM, CaCl2 3.5 mM, pH 6.9) the surface activity was assessed in a pulsating bubble surfactometer at a final phospholipid concentration of 1 mg/ml [19]. In addition, the biophysical activity was assessed in a capillary surfactometer [20]. Briefly, glass capillaries (0.255 mm internal diameter at its most narrow part) were filled with 0.5 μl of the above described surfactant suspension and assessed for the capability at 37°C to keep the capillary open over a time period of 120 seconds. Water (0% of the time open) and a purified bovine surfactant (96.7% of the time open) served as a control [20].
Reliability and precision of the measurements of surface activity was assured by the regular assessment of a well functioning bovine surfactant, water and methanol each working day (Fig. 1). Appropriate and stable readings were obtained over a period of more than 2 years for the pulsating bubble surfactometer and also the capillary surfactometer, except for two days with high readings in the capillary surfactometer towards the end of the study (Fig. 1, lower panel). These high readings were related to problems with the diameter of the capillaries and new batches, which gave the expected lower values, were used.
Figure 1 Quality control assurance by the assessment the function of a bovine surfactant (minimal surface tension, closed squares; surface tension after adsorption, open squares) and of water (minimal surface tension, closed circles; surface tension after adsorption, open squares) in the pulsating bubble surfactometer (upper panel) and the capillary surfactometer (lower panel) (bovine surfactant, closed squares, water, open squares) over a period of more than two years.
Statistical analysis
The nonparametric ANOVA, i.e. Friedman test was used for comparison of repeated measured values over the study period at the 3 time points, followed by the Dunn's post-hoc test to detect differences between each time points. Correlations were calculated with the Spearman rank test. Non-paired comparisons were made with the nonparametric Mann-Whitney test. The results are given as medians and ranges or as individual values (mean of 2 to 4 determinations). A p level of <0.05 was considered statistically significant. Statistical analysis was done with Prism 4.
Results
Minimum surface tension increased, reflecting deteriorating function, over the 3-year period (Fig. 2, left upper panel). This was paralleled by an increase in the relative and total number of neutrophils (Fig. 2, right panels). The increase of minimum surface tension over the 3 years was directly proportional to an increase in neutrophils (Fig. 3, upper panel). The total amount of phospholipids in BAL fluid was higher in the initial BAL than in the two following studies (Tab. 1). This finding is not responsible for the observed loss of surfactant function, which was assessed at a fixed phospholipid concentration. Surfactant function assessed in the capillary surfactometer, which was already severely reduced at baseline compared to values in healthy children, improved slightly after 1.5 years, but did not change significantly over the 3 year period (Fig. 2).
Figure 2 Surface activity of a surfactant fraction obtained from bronchoalveolar lavages in 20 patients with cystic fibrosis assessed in a pulsating bubble surfactometer (upper left panel), i.e. expressed as surface tension at minimum bubble radius (squares) and surface tension at adsorption (triangles), assessed in a capillary surfactometer (lower left panel), i.e. expressed as the percentage of time open of a small capillary of total time. The percentage of neutrophils (upper right panel) and the absolute number of neutrophils (lower right panel) as markers of inflammation in their BAL fluids are also indicated. Data are given as mean and standard error of the mean. Differences between the different time points were assessed by nonparametric Friedman ANOVA, followed by Dunn's post hoc multiple comparison test. The asterisks indicates a significantly different value, compared to the first data point (* P < 0.05, ** P < 0.01). The open symbols represent values in normal subjects assessed with the same methodolgy [7,29].
Figure 3 Correlation of the change in surfactant activity, assessed as minimal surface tension, to the change of the fraction of polymorphonuclear leukocytes in BAL (upper panel) and to the change of small airway function, assessed as MEF75/25 (lower panel). Changes were calculated as the difference between the last visit at 3 years and the first visit at start of the study. Spearman rank correlations and linear regression line are indicated.
Table 1 Surfactant proteins (SP-) A, B, C and D, total proteins and phospholipids in three consecutive bronchoalveolar lavages (BAL) in patients with cystic fibrosis over a 3 year period
1. BAL at start of the study 2. BAL after 1.5 years 3. BAL after 3 years
SP-A (ng/ml) 5312.0 (2409.0 – 13218.0) 5580.0 (1334.0 – 13048.0) 3782.0 (128.6 – 13624.0)
SP-B (ng/ml) 397.5 (247.2 – 1219.0) 897.5 (472.5 – 1252.0)† 847.0 (375.5 – 1602.0) **
SP-C (ng/ml) 635.0 (37.0 – 1557.0) 638.7 (391.5 – 2012.0) 581.2 (15.5 – 1489.0)
SP-D (ng/ml) 7.41 (1.17 – 22.06) 5.41 (1.48 – 43.46) 6.25 (0.00 – 48.21)
Protein (μg/ml) 104.70 (52.60 – 382.40) 97.88 (46.60 – 193.40) 94.20 (38.44 – 258.80)
Phospholipids (μg/ml) 50.37 (19.45 – 102.50) 33.56 (20.75 – 429.40) 32.64 (14.84 – 70.00) **
Comparisons were made by Friedman test followed by Dunn's post hoc test. Differences between the beginning and the end of the study are indicated by **, differences between the first and second BAL are indicated by †. Level of significance († P < 0.05, ** P < 0.01). Data are median (range) of 20 patients.
Lung function assessed as FEV1 significantly deteriorated over time (Fig. 4, upper panel). This trend was even more pronounced for lung function parameters representing predominantly the small airways (MEF75/25 and MEF25) (Fig. 4, lower panels). The changes of all three measures of airway obstruction used here, were negatively correlated to an increase of minimal surface tension (FEV1 r = -0.560, P = 0.009, MEF25%VC r = -0.536, P = 0.015 and MEF75/25%VC see Fig. 3, lower panel).
Figure 4 Lung function over time in 20 subjects with CF. FEV1, MEF72/25%VC and MEF25%VC were significantly lower 3 years after start of the study (Friedman ANOVA, followed by Dunn' post hoc multiple comparison test, * p < 0.05, *** p < 0.001). FVC did not change.
The concentration of the surfactant specific proteins SP-A, C and D, as well as the total protein content of BAL fluid did not change significantly, whereas SP-B was higher after 1.5 and 3 years than at the beginning (Table 1).
In the absence of reduced concentrations of the surfactant proteins, proteolytic attack resulting in inactivation and the presence of fragments of the proteins, may also be responsible for dysfunction of surfactant [21,22]. No such fragments were found by Western blotting for SP-B and SP-C in 21 CF patients over the range of neutrophilic inflammation, present in this cohort (Tafel, Latzin and Griese, unpublished observations). BALs from 15 CF patients from this cohort were screened for proteolytic fragments of SP-A. In 3 subjects weak bands were found; one band at 23 kD, one at 18 kD and in another subject one at 16 kD. These bands contributed to less than 1% of the total SP-A present on each of the blots (Wassilewa and Griese, unpublished data).
The surface active fraction of pulmonary surfactant, i.e. the large surfactant aggregates (LA) can be separated from the less active small aggregates, i.e. SA, which may also inhibit surfactant function. We next investigated the contribution of the SA fraction to overall surfactant function. SA added to LA at the original ratio present in bronchoalveolar lavage, significantly reduced surfactant function both, in the pulsating and in the capillary surfactometer (Fig. 5, left panels). Interestingly, this effect was not related to the protein content of SA (Fig. 5, right panels). To clarify this issue further, we asked the question, if the SA protein which has a similar protein composition as serum [21], differed from serum in its potency to inhibit surfactant, i.e. from the presence of components that have an intrinsically higher capacity to inhibit surfactant function. No such differences were found at two representative concentrations of SA investigated (Tab. 2).
Figure 5 Surfactant activity assessed in the pulsating bubble surfactometer (upper panels) and the capillary surfactometer (lower panels). The large aggregate (LA) surfactant fraction was recombined with the appropriate small aggregates (SA) of surfactant, demonstrating significant inhibitory activity (left side panels). This inhibitory activity of SA was not dependent on its amount, i.e. was not the sole determinate of surface activity (right side panels). Comparisons by Mann-Whitney test and correlations by Spearman rank test.
Table 2 Effect of small aggregates (SA) or serum on the surface activity of a well functioning proving surfactant, assessed in a pulsating bubble surfactometer and in a capillary surfactometer
Pulsating bubble surfactometer P Capillary surfactometer P
Minimal surface tension (mN/m) % open
Bovine surfactant 2.3 (0.0 – 3.50) 5 100.0 (99.9 – 100.0) 5
+ 1.5 g/l SA protein 2.3 (1.5 – 6.6) 5 ns 3.35 (0.2 – 10.3) 5 ns
+ 1.5 g/l serum protein 1.9 (1.1 – 8.6) 6 0.95 (0.1 – 8.5) 6
+ 4.5 g/l SA protein 8.6 (2.3 – 24.6) 5 ns 1.03 (0.6 – 2.3) 5 ns
+ 4.5 g/l serum protein 19.6 (1.5 – 31.9) 7 4.83 (0.3 – 35.5) 7
Data are medians (range) from n determination. Comparisons between the effects of the addition of serum and SA protein were done by Wilcoxon test.
Discussion
This is the first study to assess surfactant function, lung function and airway inflammation longitudinally in CF patients. The results demonstrate loss of surfactant function over time parallel to an increase in neutrophilic inflammation and a decrease in small airway function. These data support the hypothesis that inflammation leads to disturbances of the local airspace environment, including properties of the surfactant system resulting in a loss of patency of small airways.
The mechanism(s) responsible for this loss of surfactant function remain to be resolved. An altered phospholipid composition could contribute to reduced surface activity, but this appears rather unlikely, as no significant deviations have previously been found in a group of younger CF patients [4,23]. Similarly, no age dependency of the four surfactant proteins, total phospholipid and total protein, was observed in the whole cohort of CF patients from the BEAT study spanning an age range from 5 to 31 years (n = 75) [7]. Here we did not find reductions in the concentrations of the four surfactant proteins over time, suggesting that the loss of surface activity cannot be easily explained by a single factor like an altered expression of one of the surfactant proteins. Alterations of the integrity and function of the surfactant proteins without a change in their concentration, such as by proteolytic degradation, were largely excluded as we found no such fragments for SP-B and SP-C (Tafel, Latzin and Griese, unpublished data). For SP-A in 3 out of 15 investigated patients very small amounts of SP-A degradation products (<1% of total SP-A present) were detected by Western blotting, which was unlikely to be of functional relevance (Wassilewa and Griese, unpublished data). In addition to proteolytic damage, oxidative modification of the proteins may occur and contribute to surfactant protein dysfunction, however this was not assessed systematically in this cohort [21,22,24,25]. Little pronounced, but continuously ongoing inflammatory processes may gradually evoke changes in the function of the surface active fraction of surfactant, which may be responsible for the observed loss of biophysical surfactant activity. The principal surface active fraction of pulmonary surfactant recovered from the airspaces is represented the large surfactant aggregates (LA). Early changes in the LA surfactant micro-texture like an increased admixture of inactivating components from the SA fraction with enhanced local inflammation could be involved [26]. This explanation does not necessitate changes in the intrinsic inhibitory potency of the SA proteins and is in accordance with our results that SA fractions did not differ in their surfactant function and had a similar inhibitory potency as serum.
In our previous cross-sectional study of the same cohort of CF patients, we observed only a marginally decreased surfactant function assessed in the pulsating bubble surfactometer [7]. When those samples were analysed in the more sensitive capillary surfactometer, which simulates a small airway, the functional capacity of the surfactant was more extensively impaired with a mean capillary openness below 20% [7]. Open small airways are critical for the reduction of overinflation, the deposition of anti-inflammatory and anti-microbial treatments in those areas of the lungs where the initial neutrophilic inflammation begins to destroy the lung's integrity of patients with CF. The lack of change over time observed in this study with the capillary surfactometer may be caused by this marked reduction observed already at baseline rendering this methodology too sensitive to detect further impairments in surfactant function over time.
Of interest is a lack of correlation between minimal surface tension assessed in the pulsating bubble surfactometer and the % open assessed in the capillary surfacometer (r = -0.060; P = 0.768), when the same samples were assessed. Also, the changes of these variables over time did not correlate with each other. Identical results were obtained previously in the larger BEAT cohort (n = 75) investigated similarly for these two variables (r = -0.140; P = 0.095) [7]. Together these data demonstrate that the two techniques have different sensitivities and may also measure two distinct aspects of surfactant activity of airway specimens.
Using the pulsating bubble surfactometer and assessing minimal surface tension, loss of surfactant activity was detected with time. The concordance of the increase of minimal surface tension, loss of lung function, including small airways function and their inverse correlation support the concept that the surfactant function may be relevant for the stability of small airways [20,27]. Thus treatment strategies directed to keep the small airways patent, i.e. physiotherapeutic interventions maintaining a positive expiratory pressure or inhalation of exogenous surfactant [28] may be relevant for the treatment of early CF lung disease.
There are some important limitations of this study that need to be taken into account when interpreting the data. First, the number of patients reported in this cohort is relatively small and consists of a subgroup of the total study cohort [16]. However, we found no significant differences in CFTR genotype, sex, baseline lung function, bacterial colonisation and BAL neutrophils at baseline between the cohort reported here and the total BEAT study population. In the original report of the BEAT study results, we have observed an inhibiting effect of rhDNase on airway inflammation over time and this may also influence the data reported here. When we compared the changes in a subgroup analysis, we did not observe any differences in change of surfactant function over time between the 2 groups, but this finding has to be interpreted with caution, because the number of patients included is too small to detect effects of rhDNase. Similarly, the other variables assessed here, i.e. surfactant function and airspace inflammation were not different between the subgroups. Finally, bacterial infection changes over time and this may impact upon both inflammation and surfactant function. Again, subgroup analysis is problematic because of small numbers. The change in surfactant function showed no relationship to changes in bacterial colonisation, which may be related to the fact that most patients were already infected with CF pathogens. The effect of bacterial infection should therefore ideally be studied in a population which does not have airway infection initially.
Our data are in agreement with previous cross-sectional studies in children with CF including patients with impaired lung function, where minimal surface tension was reduced, whereas the concentrations of total phospholipids and of dipalmitoyl phosphatidylcholine were within the normal range [4,23]. Similar observations were made in older children and young adults [19], suggesting that there may be an early, but secondary, derangement of the biophysical surfactant properties in CF. With this longitudinal study of CF patients we hoped to eliminate a substantial fraction of the variability by investigating the changes within individuals overtime. Despite a large cohort of patients to start with [16], eventually only 20 complete triplets of sufficient BAL fluid were available for all the biophysical measurements. Nevertheless, we found a clear increase in minimal surface tension over the time period investigated and also correlations between these changes, the changes in lung function and an increase in neutrophils.
In summary, pulmonary surfactant function, lung function and airway inflammation were assessed sequentially over a 3-year period in CF patients with initial lung function within the normal range. An increase in neutrophil inflammation over time was associated with a decrease in small airway function and a loss of surfactant activity. These findings highlight the importance of early changes in small airway biology and strongly suggest that specific therapeutic interventions targeting this region of the lung are warranted.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MG designed the experiments on the BAL samples collected in the clinical study, drafted the manuscript. RE carried out the biophysical and the biochemical measurements, did the data calculations, performed the statistical analysis and participated in the drafting of the manuscript. RS carried out the SP-B and SP-C immunoassays. FR together with MG, MB, ER and KP conceived the study, participated in its design, coordination, practical realization and helped to draft the manuscript. All authors read and approved the final manuscript.
Acknowledgements
We thank all the staff in the clinics and in the laboratories at the different participating centers for their excellent collaboration in this study. This study was supported by a grant from the German CF foundation (Mukoviszidose e. V.), Hoffmann-La Roche, Germany and a grant of the DFG (Gr 970/7-1). The manuscript contains parts of the medical thesis of Robert Essl.
==== Refs
Gibson RL Burns JL Ramsey BW Pathophysiology and management of pulmonary infections in cystic fibrosis Am J Resp Crit Care Med 2003 168 918 951 14555458 10.1164/rccm.200304-505SO
Enhorning G Pulmonary surfactant function in alveoli and conducting airways Can Respir J 1996 3 21 27
Griese M Pulmonary surfactant in health and lung diseases: state of the art Eur Respir J 1999 13 1455 1476 10445627
Mander A Langton-Hewer S Bernhard W Warner JO Postle AD Altered phospholipid composition and aggregate structure of lung surfactant is associated with impaired lung function in young children with respiratory infections Am J Respir Cell Mol Biol 2002 27 714 721 12444031
Noah TL Murphy PC Alink JJ Leigh MW Hull WM Stahlman MT J.A. W Bronchoalveolar lavage fluid SP-A and SP-D are inversely related to inflammation in early cystic fibrosis Am J Respir Crit Care Med 2003 168 685 691 12829455 10.1164/rccm.200301-005OC
Hull J South M Phelan P Grimwood K Surfactant composition in infants and young children with cystic fibrosis Am J Respir Crit Care Med 1997 156 161 165 9230741
Griese M Essl R Schmidt R Rietschel E Ratjen F Ballmann M Paul K Pulmonary surfactant, lung function and endobronchial inflammation in cystic fibrosis Am J Resp Crit Care Med 2004 170 1000 1005 15271694 10.1164/rccm.200405-575OC
Gustafsson PM Aurora P Lindblad A Evaluation of ventilation maldistribution as an early indicator of lung disease in children with cystic fibrosis Eur Respir J 2003 22 972 979 14680088
Landau LI Mellis C Phelan P Bristowe B Mc Lennan L Small airway disease in children: no test is the best Thorax 1979 34 217 223 483188
de Jong PA Nakano Y Lequin MH Mayo JR Woods R Pare PD Tiddens HA Progressive damage on high resolution computed tomography despite stable lung function in cystic fibrosis Eur Respir J 2004 23 93 97 14738238 10.1183/09031936.03.00006603
Robinson TE Leung AN Northway WH Blankenberg FG Chan FP Bloch DA Holmes TH Moss RB Composite spirometric-computed tomography outcome measure in early cystic fibrosis lung disease Am J Respir Crit Care Med 2003 168 588 593 12746252 10.1164/rccm.200209-1093OC
Armstrong DS Grimwood K Carlin JB Carzino R Gutièrrez JP Hull J Olinsky A Phelan EM Robertson CF Phelan PD Lower airway inflammation in infants and young children with cystic fibrosis Am J Respir Crit Care Med 1997 156 1197 1204 9351622
Khan TZ Wagener JS Bost T Martinez J Accurso FJ Riches DWH Early pulmonary inflammation in infants with cystic fibrosis Am J Respir Crit Care Med 1995 151 1075 1082 7697234
Griese M Koch N Paul K Rietschel E Ballmann M Ratjen F Schams A Beck J Biophysical activity of pulmonary surfactant in cystic fibrosis Pediatr Pulmonol 2000 30 297 297 11015129 10.1002/1099-0496(200011)30:5<393::AID-PPUL5>3.0.CO;2-W
Ratjen F Rietschel E Griese M Ballmann M Kleinau I Döring G Reinhardt D Paul K for the bronchoalveolar lavage for the evaluation of anti-inflammatory treatment (BEAT) study group Fractional analysis of bronchoalveolar lavage fluid cytology in cystic fibrosis patients with normal lung function Eur Respir J 2000 15 141 145 10678636
Paul K Rietschel E Ballmann M Griese M Worlitzsch D Shute J Chen C Schink T Doring G van Konigsbruggen S Wahn U Ratjen F Effect of treatment with dornase alpha on airway inflammation in cystic fibrosis patients Am J Respir Crit Care Med 2004 169 719 725 14684561 10.1164/rccm.200307-959OC
Griese M Maderlechner N Ahrens P Kitz R Surfactant proteins A and D in children with pulmonary disease due to gastroesophageal reflux Am J Resp Crit Care Med 2002 165 1546 1550 12045131 10.1164/rccm.2107147
Schmidt R Steinhilber W Ruppert C Grimminger F Seeger W Günther A An ELISA technique for quantification of surfactant apoprotein (SP)-C in bronchoalveolar lavage fluid Am J Respir Crit Care Med 2002 165 470 474 11850338
Griese M Birrer P Demirsoy A Pulmonary surfactant in cystic fibrosis Eur Respir J 1997 10 1983 1988 9311489 10.1183/09031936.97.10091983
Enhorning G Holm BA Disruption of pulmonary surfactant´s ability to maintain openness of a narrow tube. J Appl Physiol 1993 74 2922 2927 8365993
von Bredow C Birrer P Griese M Surfactant protein A and other bronchoalveolar lavage proteins are altered in cystic fibrosis Eur Respir J 2001 17 716 722 11401069 10.1183/09031936.01.17407160
Griese M Wiesener A Lottspeich F v.Bredow C Limited proteolysis of surfactant protein D causes a loss of its calcium-dependent lectin functions Biochim Biophys Acta 2003 1638 157 163 12853121
Postle AD Mander A Reid KBM Wang JY Wright SM Moustaki M Warner JO Deficient hydrophilic lung surfactant proteins A und D with normal surfactant phospholipid molecular species in cystic fibrosis Am J Resp Cell Mol Biol 1999 20 90 98
Hirche TO Crouch EC Espinola M Brokelman TJ Mecham RP DeSilva N Cooley J Remold-O'Donnell E Belaaouaj A Neutrophil serine proteinases inactivate surfactant protein D by cleaving within a conserved subregion of the carbohydrate recognition domain J Biol Chem 2004 279 27688 27698 15078883 10.1074/jbc.M402936200
Malloy JL Veldhuizen RA Thibodeaux BA O'callaghan RJ Wright JR Pseudomonas aeruginosa protease IV degrades surfactant proteins and inhibits surfactant host defense and biophysical functions Am J Physiol Lung Cell Mol Physiol 2004 15516485
Seeger W Grube C Günther A Schmidt R Surfactant inhibition by plasma proteins: differential sensitivity of various surfactant preparations Eur Respir J 1993 6 971 977 8370446
Macklem PT Proctor DF Hogg JC The stability of peripheral airways Respir Physiol 1970 8 191 203 5413420 10.1016/0034-5687(70)90015-0
Griese M Bufler P Teller J Reinhardt D Nebulization of a bovine surfactant in cystic fibrosis: a pilot study Eur Respir J 1997 10 1989 1997 9311490 10.1183/09031936.97.10091989
Griese M Felber J Reiter K Strong P Reid K Belohradsky BH Jäger G Nicolai T Airway inflammation in children with tracheostomy Pediatr Pulmonol 2004 37 356 361 15022133 10.1002/ppul.10432
|
16274485
|
PMC1308867
|
CC BY
|
2021-01-04 16:36:25
|
no
|
Respir Res. 2005 Nov 7; 6(1):133
|
utf-8
|
Respir Res
| 2,005 |
10.1186/1465-9921-6-133
|
oa_comm
|
==== Front
Respir ResRespiratory Research1465-99211465-993XBioMed Central London 1465-9921-6-1371628750910.1186/1465-9921-6-137ResearchCharacterization of lymphocyte populations in nonspecific interstitial pneumonia* Keogh Karina A [email protected] Andrew H [email protected] Thoracic Diseases Research Unit, Division of Pulmonary & Critical Care Medicine, Department of Internal Medicine, Mayo Clinic College of Medicine, Rochester. MN, 55905, USA2005 15 11 2005 6 1 137 137 11 3 2005 15 11 2005 Copyright © 2005 Keogh and Limper; licensee BioMed Central Ltd.2005Keogh and Limper; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Study objectives
Nonspecific interstitial pneumonia (NSIP) has been identified as a distinct entity with a more favorable prognosis and better response to immunosuppressive therapies than usual interstitial pneumonia (UIP). However the inflammatory profile of NSIP has not been characterized.
Design
Using immunohistochemistry techniques on open lung biopsy specimens, the infiltrate in NSIP was characterized in terms of T and B cells, and macrophages, and the T cell population further identified as either CD4 (helper) or CD8 (suppressor-cytotoxic) T cells. The extent of Th1 and Th2 cytokine producing cells was determined and compared to specimens from patients with UIP.
Results
In ten NSIP tissue samples 41.4 ± 4% of mononuclear cells expressed CD3, 24.7 ± 1.8% CD4, 19.1 ± 2% CD8, 27.4 ± 3.9% CD20, and 14.3 ± 1.6% had CD68 expression. Mononuclear cells expressed INFγ 21.9 ± 1.9% of the time and IL-4 in 3.0 ± 1%. In contrast, biopsies from eight patients with UIP demonstrated substantially less cellular staining for either cytokine (INFγ; 4.6 ± 1.7% and IL-4; 0.6 ± 0.3%). Significant populations of CD20 positive B-cells were also identified.
Conclusion
The lymphocytic infiltrate in NSIP is characterized by an elevated CD4/CD8 T-cell ratio, and is predominantly of Th1 type, with additional populations rich in B-cells. Such features are consistent with the favorable clinical course observed in patients with NSIP compared to UIP.
CytokinesLymphocytesNonspecific interstitial pneumonitisPulmonary fibrosisUsual interstitial pneumonitis
==== Body
Introduction
Nonspecific interstitial pneumonia (NSIP) has recently been identified as a distinct form of idiopathic interstitial pneumonia, distinguishable from usual interstitial pneumonia (UIP). NSIP has been associated with better response to immunosuppressive therapies and a more favorable prognosis [1-4]. Histological examination demonstrates that NSIP is characterized by a mononuclear lymphocytic interstitial infiltrate, with occasional foci of fibroblasts and variable collagen deposition [3,5]. However, the prevalence of B and T cell populations in NSIP, and specifically the CD4 or CD8 T cell content has not been fully defined in this disorder. Moreover, the relative Th1 or Th2 cytokine expression associated with this disease is also not yet known.
Inflammatory responses are generally categorized into two major types on the basis of the predominant cytokines secreted. Most autoimmune diseases, including pulmonary diseases such as sarcoidosis, follow a Th1 pattern, whereas allergic diseases such as asthma generally demonstrate a Th2 pattern [6,7]. The relevance of patterned cytokine expression during pulmonary fibrosis has been supported by a number of studies [8-11]. For instance, Th1 cells produce predominantly interferon gamma (IFNγ) and interleukin 2 (IL-2), which impair fibroblast activation and proliferation and suppress collagen production. In contrast, Th2 cells secrete IL-4, IL-10, and IL-13. Th2 cells may thereby act to stimulate fibroblast growth and promote collagen production. Thus, the relative extent of Th1 and Th2 cytokine production may underlie the tendency of various interstitial lung diseases toward more or less rapid progression, and may further limit the extent of reversibility in these disorders.
Accordingly, the following study was performed to determine the cellular populations present in lung tissue from patients with NSIP. We first characterized the infiltrate in NSIP in terms of T and B cells, and macrophages, and further identified the T cell population as either CD4 (helper) or CD8 (suppressor-cytotoxic) T cells. This was undertaken utilizing immunohistochemistry on tissues obtained by open lung biopsy. As a second aim, we determined the extent of Th1 and Th2 cytokine producing cells in lung tissues obtained from these patients with NSIP. In comparison lung tissues from patients with UIP were analyzed concurrently.
Materials and methods
Subjects and Tissue Collection
The Mayo Foundation Institutional Review Board approved these studies. All biopsies were obtained during the routine clinical care of these patients. The study population consisted of ten patients with a pathologic confirmation of NSIP established by experienced pulmonary pathologists. The diagnosis was based on histological findings in biopsies obtained by video assisted thoracoscopy, between November 1997 and present according to previously published criteria [3,12]. There were seven women and three men in our study population, with a mean age of 50.5 years (range 17–66) (Table 1). Three were prior smokers and seven were never smokers. All except one patient, had moderate to severe inflammation present on the biopsy. In case number 1, the inflammation was judged as mild to moderate. One patient was receiving high dose intravenous methylprednisolone at the time of biopsy (case 5), and two were receiving oral prednisone. Three additional patients had received prednisone within the six months prior to biopsy (Table 1). In preliminary studies, tonsillar tissue was used as a lymphocyte-rich control to confirm that the immunohistochemical procedures employed were robust in their ability to detect the specified cellular antigens. For comparison of Th1 (IFNγ and Th2 (IL-4) cytokine expression, samples from a group of patients with UIP were studied in parallel. This population consisted of eight patients with a clinicopathologic diagnosis of UIP and a mean age of 65.6 years (range 60–81). Four were prior smokers and four had never smoked. These UIP patients were previously reported as part of a separate, strictly morphological, study [1].
Table 1 NSIP Patient Characteristics
Case Age Sex Background Illnesses Steroids prior to Biopsy CD4/CD8 Lymphoid Aggregates* INFγ % IL-4 %
1 50 F - - 1.9 - 24.5 1
2 66 F - Previous 1.1 ++ 19.8 8.2
3 66 M - Current 1.5 - 14 3.9
4 51 M - - 1.1 + 17.3 -
5 17 F ARF† Current 2.1 ++ 20.7 4.8
6 45 F Current 0.8 + 12.8 1.1
7 65 F Nitrofurantoin exposure Previous 0.9 ++ 26 0.5
8 57 F - - 1.5 ++ 25.5 2.4
9 31 M - Previous 1.5 ++ 28.1 7.2
10 57 F CREST - 1.2 ++ 30 0.7
• Lymphoid aggregates, + = small aggregates, ++ large aggregates, - more fibrotic
• † = acute respiratory failure present clinically prior to biopsy.
Immunohistochemical Evaluation
Formalin-fixed, paraffin embedded sections of 5-μm thickness were deparaffinized through three, 20 minute, exchanges of xylene. The tissues were then rehydrated using a graded series of alcohol washes (100%, 100%, 95%, 70%, 50% and 30%), and then incubated for 30 minutes in 0.5% hydrogen peroxide to quench endogenous peroxidase activity. After 30-minute incubation with blocking serum (1% horse serum for mouse primary antibodies and 1% goat serum for rabbit antibodies), the primary antibodies were applied. All primary antibodies were mouse monoclonal antibodies, with the exception of a CD3 rabbit polyclonal antibody. The primary antibodies evaluated were those recognizing CD3 (5 μg/ml, DAKO Corporation, Carpinteria, CA), CD4 (41 μg/ml, Novocastra Laboratories, Newcastle upon Tyne, UK), CD8 (1 μg/ml Serotec, Raleigh, NC), CD20 (8 μg/ml, DAKO) a B cell marker, CD68 (used undiluted, DAKO) a monocyte/macrophage marker, IL-4 (15 μg/ml, Stem Cell Technologies, Vancouver, BC), and INFγ (2.5 μg/ml, Stem Cell Technologies) [13]. Antibody dilutions were applied uniformly in parallel across all tissues studied. Enzymatic pretreatment for antigen retrieval was necessary for the detection of CD3 and INFγ, using Proteinase K (20 μg/ml for 10 minutes at room temperature, Invitrogen Corporation, Carlsbad, CA), and IL-4, using Protease XXV (1000 μg/ml for 10 minutes at 37°C, NeoMarkers, Fremont, CA). Heat retrieval of epitopes by boiling was used for the CD4 and CD8 studies in the presence of 1 mM EDTA; pH 8 for 10 minutes (Sigma, St Louis, MO), and in the case of CD20, utilizing 10 mM sodium citrate buffer; pH 6 for 10 minutes (Sigma). Primary antibody binding was detected using the avidin-biotin immunoperoxidase method (Vectastain Elite ABC Kit, Vector Laboratories, Burlingame, CA) with 3-amino-9-ethyl-carbazole substrate (AEC) as the colorimetric substrate, producing a red to brown pigment. The sections were counterstained with 1% hematoxylin. The percentage of positively stained cells in each sample was determined by counting stained and non-stained mononuclear cells in 5 randomly selected contiguous high-power fields (400× magnification). Fibroblasts, epithelial, endothelial cells and intravascular cells were excluded in the enumeration procedure [14]. The coefficient of variation (standard deviation/mean × 100%) of the enumeration procedure was ~17% on repeated counting of the same stained sections.
Statistical analysis
Descriptive analyses were performed using the statistical software package, JMP version 4.0 (SAS Institute Inc., NC). Results are expressed as mean ± standard error of the mean. Differences between non-parametric groups were analyzed using Wilcoxon/Kruskal-Wallis tests. P < 0.05 was considered a statistically significant difference. Coefficients of variation were calculated from triplicate slide counts from nine slides recounted in a random blinded manner.
Results
The NSIP tissues were rich in mononuclear cells, with a relatively high CD4/CD8 ratio, and a large number of B cells. Abundant lymphoid aggregates were seen in eight of ten specimens. Overall 41.4 ± 4.0% of interstitial mononuclear cells expressed the pan T cell surface marker, CD3 (Figure 1). These cells were found scattered throughout the interstitium and at the peripheries of lymphoid follicles. In addition 24.7 ± 1.8% of the total mononuclear cells were classified as CD4 lymphocytes (59.7% of the T cells). CD4 cells were found primarily in circumferential mantles around lymphoid follicles, but were also scattered throughout the interstitial spaces. There was also low grade staining of additional cells with this antibody, which appeared morphologically to represent alveolar macrophages. Furthermore, 19.1 ± 2.0% of the total mononuclear cell population were classified as CD8 lymphocytes (46.5% of the T cells) (Figure 1). CD8 expressing cells were found as strongly stained individual cells scattered throughout the interstitium. The CD4/CD8 ratio was 1.36 ± 0.13 (Table 1). When values from patients on steroid treatment at the time of biopsy were excluded the ratio was not found to be significantly different. In addition, the B cell marker CD20 was present on 27.4 ± 3.9% of the mononuclear cells. B cells were detected as strongly stained cells primarily within lymphoid follicles. We further observed that 14.3 ± 1.6% of mononuclear cells displayed the macrophage/monocyte marker CD68. These were localized as clusters in the airspaces and as occasional individual cells within the interstitium (Figure 1). When intra-alveolar cells were included 30.0 +/- 3.1% of all mononuclear cells expressed the CD68 antigen.
Figure 1 Immunohistochemical localization of CD3, CD4, CD8, CD20 and CD68 in lung tissue with NSIP. Bound primary antibodies to these antigens were detected by an avidin-biotin immunoperoxidase method, with AEC substrate (arrow) and counterstained with 1% hematoxylin. Staining for CD3 was present on lymphocytes located throughout the interstitium and lymphoid follicles. CD4 cells were primarily present in a perifollicular location. CD8 expressing cells were scattered throughout the interstitium. CD20 was also primarily localized to the follicles. The CD68 macrophage/monocyte marker was found in clusters of the intra-alveolar spaces and on occasional individual cells within the interstitium. A. Percentage of cells stained with each individual antibody, expressed as a percentage of mononuclear cells within the interstitium. Values are reported as mean ± SEM (N = 10 patients).
Cytokine expression was substantially greater in NSIP compared with UIP tissues. INFγ is a prototypic cytokine characteristic of a Th1 type cytokine response. In these tissues from with NSIP, cytoplasmic staining for IFNγ was demonstrated in 21.9 ± 1.9% of the interstitial mononuclear cells (Figure 2). There was also abundant staining of intra-alveolar macrophages and some staining of epithelial cells for cell associated IFNγ. In contrast, IL-4 expression is more typical of a Th2 patterned cytokine response. IL-4 was detected in 3.0 ± 1.0% of mononuclear cells in subjects with NSIP. These IL-4 expressing cells were found scattered throughout the biopsies as individual mononuclear cells. There was also some limited background staining of epithelial cells for IL-4. For comparison, 9 UIP samples were also evaluated for IFNγ and IL-4 expression. There was much less staining for either cytokine in the UIP samples (Figure 2). In the UIP biopsies, very occasional mononuclear cells (4.6 ± 1.7%) and only occasional epithelial cells displayed any cell associated INFγ. Virtually no cells exhibited IL-4 (0.6 ± 0.3%) in the UIP tissues evaluated.
Figure 2 Immunohistochemical localization of cell-associated INFγ and IL-4 in NSIP and UIP lung biopsies. NSIP tissue exhibited diffuse expression of INFγ on lymphocytes. INFγ was also associated with macrophages and epithelial cells. Few cells stained positive for IL-4 in NSIP lung. Both stains were also significantly less prominent in UIP. The percentage of stained cells was enumerated for the NISP and UIP samples. Values are reported as mean ± SEM for 10 NSIP and 8 UIP biopsies. (* Denotes significant differences in the extent of cellular staining, for INFγ P = 0.0005, for IL-4 p = 0.045 comparing NSIP and UIP by Wilcoxon/Kruskal-Wallis Tests).
Discussion
Alveolar interstitial lymphocytes are rare in normal lung parenchyma. The presence of interstitial and alveolar lymphocytes in NSIP has been previously documented both on histology and bronchoalveolar lavage (BAL) [3,4,14,15]. This study was undertaken to further characterize the inflammatory cell infiltrate in NSIP. The key findings were; 1) The cellular infiltrate in NSIP is largely composed of lymphocytes with a relatively high CD4/CD8 ratio, 2) A large proportion of the mononuclear cells express the B cell specific antigen CD20, and 3) Cytokine expression was substantially greater in NSIP compared with UIP tissues. This cytokine expression in NSIP was predominantly a Th1 patterned response.
The observed CD4/CD8 cellular ratio of 1.36 ± 0.13 was higher than expected and is higher than has been reported in two prior studies [4,17]. One of these previous investigations evaluated only BAL data, and the other studied both BAL and histology in NSIP and pulmonary fibrosis associated with connective tissue disease. The perifollicular locations of these cells make them less likely to be washed from the alveoli during BAL [16,17]. In addition, the majority of our NSIP biopsies demonstrated a high degree of cellularity rather than fibrosis, which may have also influenced our observations. Our finding that this cellularity contains a large number of CD20 positive B cells is of interest, and may suggest a new target for treatment of NSIP. In patients who are not responding to traditional therapy, there may be a possible role for agents such as rituximab, a monoclonal antibody against CD20 [18-20].
Clinically, NSIP behaves as a more inflammatory process, with greater responsiveness to immunosuppressive therapy, in distinct contrast to UIP. Our findings of cytokine-rich infiltrates in NSIP further support these observations. In our study, the cytokine expression pattern in NSIP appears to be consistent with a predominant Th1 response. The dichotomy between Th1 and Th2 cells was first demonstrated in murine CD4 T cell clones [21]. It has since been identified in humans with chronic inflammatory lesions [22-24].
IFNγ, which is secreted from Th1 cells, has been shown in previous studies to suppress fibroblast activity in vitro and in murine models of bleomycin induced fibrosis [6,24-27]. Investigations have also suggested that patients with UIP have impaired production of INFγ and that the administration of IFNγ can alter their disease process [10,28]. In contrast, Th2 cells secrete IL-4, which has been implicated as a fibroblast-stimulating agent [29,30]. IL-4 has been found to be upregulated in some murine models of fibrosis [31]. Enhanced production of IL-4 has also been observed in pulmonary fibrosis associated with systemic sclerosis, which also exhibits a lower Th2/Th1 ratio than UIP, and further is associated with a substantially higher level of INFγ production in tissue [9]. The level of cell staining for IL-4 in the UIP samples in our study was somewhat lower than has previously been reported. This may have been influenced by the general lack of cellularity of our UIP samples, as UIP is associated with heterogeneous involvement of the lung parenchyma. These previous studies on UIP suggested a Th2 type response with very low levels of INFγ, and higher levels of IL-4 [9,10].
Our study indicates that there is a substantial increase in IFNγ production in NSIP when compared to both our UIP specimens and previous publications [9,10]. This increased level of IFNγ in NSIP, would be expected to counteract the postulated pro-fibrotic effect of IL-4 and may help explain the relative lack of fibrotic foci in this form of idiopathic interstitial pneumonia.
One limitation to our study is the inclusion of patients in the study group who were receiving glucocorticoids, or had been treated with them in the past. This was necessary as during this time period, at our institution, only a small number of patients with interstitial lung disease were proceeding to surgical lung biopsy without a previous trial of glucocorticoids. In analyzing the results, there was no significant difference between cytokine levels based on current or previous treatment.
In conclusion, we observed that NSIP is characterized by a largely lymphocytic infiltrate, with a high CD4/CD8 ratio, rich in cytokines, predominantly exhibiting a Th1 type response. These findings may in part explain why NSIP, in comparison to UIP, follows a slower, less fibrotic course, and is more responsive to immune modulatory therapies.
Abbreviation list
AEC = 3-amino-9-ethyl-carbazole substrate
BAL = bronchoalveolar lavage
IFNγ = interferon gamma
IL = interleukin
NSIP = Nonspecific interstitial pneumonia
UIP = usual interstitial pneumonia
Note
*This work was supported by funds from the Robert N. Brewer Family Foundation, and funds
from the Mayo Foundation.
Acknowledgements
The authors thank Dr. Jay Ryu and the members of the Mayo Interstitial Lung Disease focus group for identification and clinical management of these patients with NSIP. The authors further appreciate the technical advice of Dr. Zvezdana Vuk-Pavlovic in establishing the immunohistochemical assays. This work was supported by funds from the Robert N. Brewer Family Foundation, and funds from the Mayo Foundation.
==== Refs
Bjoraker JA Ryu JH Edwin MK Prognostic significance of histopathologic subsets in idiopathic pulmonary fibrosis Am J Respir Crit Care Med 1988 157 199 203 9445300
Daniil ZD Gilchrist FC Nicholson AG A histologic pattern of nonspecific interstitial pneumonia is associated with a better prognosis than usual interstitial pneumonia in patients with cryptogenic fibrosing alveolitis Am J Respir Crit Care Med 1999 160 899 905 10471616
Katzenstein AL Fiorelli RF Nonspecific interstitial pneumonia/fibrosis. Histologic features and clinical significance Am J Surg Pathol 1994 18 136 147 8291652
Nagai S Kitaichi M Itoh H Idiopathic nonspecific interstitial pneumonia/fibrosis: comparison with idiopathic pulmonary fibrosis and BOOP Eur Respir J 1998 12 1010 1009 9863989 10.1183/09031936.98.12051010
Katzenstein AL Myers JL Nonspecific interstitial pneumonia and the other idiopathic interstitial pneumonias: classification and diagnostic criteria Am J Surg Pathol 2000 24 1 3 10632482 10.1097/00000478-200001000-00001
Agostini C Trentin L Perin A Regulation of alveolar macrophage-T cell interactions during Th1-type sarcoid inflammatory process Am J Physiol Lung Cell Mol Physiol 1999 277 240 250
Nonaka M Nonaka R Woolley K Distinct immunohistochemical localization of IL-4 in human inflamed airway tissues J Immunol 1995 155 3234 3244 7673736
Gurujeyalakshmi G Giri SN Molecular mechanisms of antifibrotic effect of interferon gamma in bleomycin-mouse model of lung fibrosis: downregulation of TGF-beta and procollagen I and III gene expression Exp Lung Res 1995 21 791 808 8556994
Majumdar S Li D Ansari T Different cytokine profiles in cryptogenic fibrosing alveolitis and fibrosing alveolitis associated with systemic sclerosis: a quantitative study of open lung biopsies Eur Respir J 1999 14 251 257 10515397 10.1034/j.1399-3003.1999.14b03.x
Wallace WA Ramage EA Lamb D A type 2 (Th2-like) pattern of immune response predominates in the pulmonary interstitium of patients with cryptogenic fibrosing alveolitis (CFA) Clin Exp Immunol 1995 101 436 441 7664490
Ziesche R Hofbauer E Wittmann K A preliminary study of long-term treatment with interferon gamma-1b and low-dose prednisolone in patients with idiopathic pulmonary fibrosis N Engl J Med 1999 341 1264 1269 10528036 10.1056/NEJM199910213411703
Myers JL NSIP, UIP, and the ABCs of idiopathic interstitial pneumonias Eur Respir J 1998 12 1003 1004 9863986 10.1183/09031936.98.12051003
Falini B Flenghi L Pileri S PG-M1: a new monoclonal antibody directed against a fixative-resistant epitope on the macrophage-restricted form of the CD68 molecule Am J Pathol 1993 142 1359 1372 7684194
Kradin RL Divertie MB Colvin RB Usual interstitial pneumonitis is a T-cell alveolitis Clin Immunol Immunopathol 1986 40 224 235 3521973 10.1016/0090-1229(86)90025-5
Fujita J Yamadori I Suemitsu I Clinical features of non-specific interstitial pneumonia Respir Med 1999 93 113 108 10464862 10.1016/S0954-6111(99)90300-1
Yamadori I Fujita J Kajitani H Lymphocyte subsets in lung tissues of interstitial pneumonia associated with untreated polymyositis/dermatomyositis Rheumatol Int 2001 21 89 93 11765227 10.1007/s00296-001-0146-y
Yamadori I Fujita J Kajitani H Lymphocyte subsets in lung tissues of non-specific interstitial pneumonia and pulmonary fibrosis associated with collagen vascular disorders: correlation with CD4/CD8 ratio in bronchoalveolar lavage Lung 2000 178 361 370 11361059 10.1007/s004080000037
Edwards JCW Cambridge G Sustained improvement in rheumatoid arthritis following a protocol designed to deplete B lymphocytes Rheumatology 2001 40 205 211 11257159 10.1093/rheumatology/40.2.205
Maloney DG Grillo-Lopez AJ White CA IDEC-C2B8 (Rituximab) Anti-CD20 monoclonal antibody therapy in patients with relapsed low-grade non-Hodgkin's lymphoma Blood 1997 90 2188 2195 9310469
Specks U Fervenza FC McDonald TJ Response of Wegener's granulomatosis to anti-CD20 chimeric monoclonal antibody therapy Arthritis Rheum 2001 44 2836 2840 11762944 10.1002/1529-0131(200112)44:12<2836::AID-ART471>3.0.CO;2-W
Mosmann TR Cherwinski H Bond MW Two types of murine helper T cell clone. I. Definition according to profiles of lymphokine activities and secreted proteins J Immunol 1986 136 2348 2357 2419430
Romagnani S Lymphokine production by human T cells in disease states Annu Rev Immunol 1994 12 227 257 8011282 10.1146/annurev.iy.12.040194.001303
Simon AK Seipelt E Sieper J Divergent T-cell cytokine patterns in inflammatory arthritis Proc Natl Acad Sci USA 1994 91 8562 8566 8078923
Duncan MR Berman B Gamma interferon is the lymphokine and beta interferon the monokine responsible for inhibition of fibroblast collagen production and late but not early fibroblast proliferation J Exp Med 1985 162 516 527 2410528 10.1084/jem.162.2.516
Elias JA Freundlich B Kern JA Cytokine networks in the regulation of inflammation and fibrosis in the lung Chest 1990 97 1439 1445 2112081
Hyde DM Henderson TS Giri SN Effect of murine gamma interferon on the cellular responses to bleomycin in mice Exp Lung Res 1998 14 687 704 2465144
Pfeffer LM Murphy JS Tamm I Interferon effects on the growth and division of human fibroblasts Exp Cell Res 1979 121 111 120 446522 10.1016/0014-4827(79)90450-6
Prior C Haslam PL In vivo levels and in vitro production of interferon-gamma in fibrosing interstitial lung diseases Clin Exp Immunol 1992 88 280 287 1572093
Sempowski GD Beckmann MP Derdak S Subsets of murine lung fibroblasts express membrane-bound and soluble IL-4 receptors. Role of IL-4 in enhancing fibroblast proliferation and collagen synthesis J Immunol 1994 152 3606 3614 7908305
Sempowski GD Derdak S Phipps RP Interleukin-4 and interferon-gamma discordantly regulate collagen biosynthesis by functionally distinct lung fibroblast subsets J Cell Physiol 1996 167 290 296 8613470 10.1002/(SICI)1097-4652(199605)167:2<290::AID-JCP13>3.0.CO;2-C
Gharaee-Kermani M Nozaki Y Hatano K Lung interleukin-4 gene expression in a murine model of bleomycin-induced pulmonary fibrosis Cytokine 2001 15 138 147 11554783 10.1006/cyto.2001.0903
|
16287509
|
PMC1308868
|
CC BY
|
2021-01-04 16:36:25
|
no
|
Respir Res. 2005 Nov 15; 6(1):137
|
utf-8
|
Respir Res
| 2,005 |
10.1186/1465-9921-6-137
|
oa_comm
|
==== Front
Thromb JThrombosis Journal1477-9560BioMed Central London 1477-9560-3-171627448310.1186/1477-9560-3-17Original Clinical InvestigationIncreased PAI-1 plasma levels and risk of death from dengue: no association with the 4G/5G promoter polymorphism Mairuhu ATA [email protected] TE [email protected] P [email protected] CE [email protected] A [email protected] SMH [email protected] Cate H [email protected] DPM [email protected] ADME [email protected] PH [email protected] Gorp ECM [email protected] Department of Internal Medicine, Slotervaart Hospital, Louwesweg 6, 1066 EC Amsterdam, The Netherlands2 Department of Paediatrics, Dr. Kariadi Hospital, Semarang, Indonesia3 Institute of Virology, Erasmus Medical Centre, P.O. Box 1738, 3000 DR Rotterdam, The Netherlands4 Department of Immunopathology, Sanquin Research, P.O. Box 9190, 1006 AD Amsterdam, The Netherlands5 Laboratory for Experimental and Clinical Immunology, Academic Medical Centre, Meibergdreef 9, 1100 DD Amsterdam, The Netherlands6 Department of Clinical Chemistry, VU Medical Centre, Amsterdam, The Netherlands7 Hematology and Clinical Chemistry Laboratory, Onze Lieve Vrouwe Gasthuis, Amsterdam, The Netherlands8 Molecular and Cytogenetics Unit, Biotechnology Laboratory, Medical Faculty Diponegoro University, Semarang, Indonesia9 Department of Internal Medicine, University Hospital Maastricht, P.O. Box 5800, 6202 AZ Maastricht, The Netherlands10 Cardiovascular Research Institute, Maastricht University, P.O. Box 616, 6200 MD Maastricht, The Netherlands11 Laboratory for Experimental Medicine, Academic Medical Centre, Meibergdreef 9, 1100 DD, Amsterdam, The Netherlands2005 7 11 2005 3 17 17 17 3 2005 7 11 2005 Copyright © 2005 Mairuhu et al; licensee BioMed Central Ltd.2005Mairuhu et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Dengue virus infected patients have high plasminogen activator inhibitor type I (PAI-1) plasma concentrations. Whether the insertion/deletion (4G/5G) polymorphism in the promotor region of the PAI-1 gene is associated with increased PAI-1 plasma concentrations and with death from dengue is unknown. We, therefore, investigated the relationship between the 4G/5G polymorphism and PAI-1 plasma concentrations in dengue patients and risk of death from dengue.
Methods
A total of 194 patients admitted to the Dr. Kariadi Hospital in Semarang, Indonesia, with clinical suspected severe dengue virus infection were enrolled. Blood samples were obtained on day of admission, days 1, 2 and 7 after admission and at a 1-month follow-up visit. Plasma concentrations of PAI-1 were measured using a sandwich ELISA kit. The PAI-1 4G/5G polymorphism was typed by allele-specific PCR analysis.
Results
Concentrations of PAI-1 on admission and peak values of PAI-1 during admission were higher than the values measured in healthy controls. Survival was significantly worse in patients with PAI-1 concentrations in the highest tertile (at admission: OR 4.7 [95% CI 0.9–23.8], peak value during admission: OR 6.3 [95%CI 1.3–30.8]). No association was found between the PAI-1 4G/5G polymorphism, and PAI-1 plasma concentrations, dengue disease severity and mortality from dengue.
Conclusion
These data suggest that the 4G/5G polymorphism has no significant influence on PAI-1 concentrations in dengue virus infected patients and is not associated with the risk of death from dengue. Other factors contributing to the variability of PAI-1 plasma concentrations in patients with dengue need to be explored.
==== Body
Background
Dengue is the most prevalent viral disease transmitted by arthropod vectors worldwide [1]. An estimated 50–100 million cases of dengue fever and 500.000 cases of dengue haemorrhagic fever resulting in around 24.000 deaths occur annually depending on epidemic activity [1,2]. At present, almost 30% of the world population is at risk for dengue virus infection and it is expected that this number will increase substantially as transmission spreads to other yet unaffected geographic regions [3]. The viruses are transmitted to humans through infected mosquitoes, and may induce clinical manifestations ranging from a mild, uncomplicated febrile illness to the more severe dengue haemorrhagic fever and dengue shock syndrome. Increased vascular permeability is thought to be central in the pathogenesis of dengue haemorrhagic fever and dengue shock syndrome, since it results in the loss of plasma from the vascular compartment, which may give rise to shock in severe cases. Bleeding manifestations are believed to result from thrombocytopenia and thrombocytopathia, but increasing evidence suggests an important role for other abnormalities in the coagulation and fibrinolytic systems [4].
Nearly all patients suffering from dengue haemorrhagic fever show some evidence of a deranged coagulation system, but coagulation abnormalities are most marked in dengue shock syndrome patients [5]. The activity of the fibrinolytic system is amongst others regulated by plasminogen activator inhibitor type I (PAI-1), of which levels may greatly increase during acute phase reactions. Indeed, levels of PAI-1 are high in particular in dengue shock syndrome patients with an adverse clinical outcome [6,7]. A single base pair insertion/deletion (4G/5G) polymorphism in the promotor region of the PAI-1 gene has been associated with increased plasma concentrations of PAI-1 and with the development of shock and death after infection with Neisseria meningitides [8,9]. We therefore investigated whether in dengue virus infected patients increased PAI-1 levels are associated with a greater risk of death from dengue and whether the 4G/4G genotype contributes to these higher levels.
Methods
Patients were selected from two observational studies conducted in the Dr. Kariadi Hospital in Semarang, Indonesia. In the first study, performed from 1996 to 1997 at the paediatric intensive care unit, a total of 50 patients with a clinical diagnosis of suspected dengue shock syndrome were studied. Blood samples were obtained from these patients on day of admission and on days 1, 2 and 7 after admission. The study protocol and the results of the measurement of PAI-1 plasma levels have been described previously [6]. The second study was performed from 2001 to 2003 at the paediatric intensive care unit and the paediatric ward. Patients, aged 3 to 14 years, admitted to the hospital with a clinical diagnosis of suspected dengue shock syndrome or with a clinical diagnosis of suspected dengue haemorrhagic fever were included. Demographic data, medical history, physical examination findings and subsequent progress for each patient were recorded on a standard data form. Blood samples were obtained on day of admission, days 1, 2 and 7 after admission and at a 1-month follow-up visit. A formal classification, according to WHO criteria [2], using all available clinical and laboratory data, was done after completion of both studies. If dengue virus infected patients did not meet the criteria for dengue haemorrhagic fever or dengue shock syndrome, they were considered to suffer from dengue fever. The controls were healthy school-aged children who had no history of dengue haemorrhagic fever or dengue shock syndrome and originated from the same geographical area as the cases.
All blood samples were centrifuged within 1–2 hours after retrieval at 15°C for 20 minutes at 1600*g. Plasma was separated, stored at -80°C and assayed batch-wise in the Netherlands after transportation on dry ice. Since coagulation test results are affected by poor sample quality, only test results of samples with no visible haemolysis or clot formation were included in this analysis. Plasma concentrations of PAI-1 were measured using two separate assays. A sandwich ELISA kit that has been described by de Boer et al [10], was used in the 1996–1997 study. In the second study a commercially available sandwich ELISA kit (Imulyse, Biopool, Sweden) was used. Genomic DNA from patients was prepared either from EDTA whole blood by means of a salting out method as described elsewhere or in case only plasma samples were available with use of the Invisorb® Spin Cell Mini Kit (Invitek GmbH, Berlin, Germany) [11]. The insertion/deletion (4G/5G) polymorphism in the promotor region of the PAI-1 gene was typed by allele-specific PCR and RFLP analysis [11]. The ethics committee of the Dr. Kariadi Hospital approved all clinical and laboratory aspects of both studies. Blood samples were taken from patients and controls provided that a parent or legal guardian gave informed consent.
Paired blood samples from both studies were tested for serologic evidence of acute dengue infection. A commercially available capture and indirect ELISA (Focus Technologies, Cypress, Calif., USA) was used for the detection of dengue virus specific IgM and IgG antibodies respectively. This was performed according to the procedures described by the manufacturer [12]. For some patients, a definitive serodiagnosis was not possible because no convalescent sample was obtained. Detection of dengue antigen and RNA was attempted in these cases using a dot blot immunoassay and a dengue serotype specific revere transcriptase PCR respectively [13]. Patients with serologic evidence of acute dengue infection, a positive dot blot and/or positive PCR were considered to have confirmed dengue virus infection. Those with definite negative serology and/or a well-substantiated alternative clinical diagnosis were classified as not dengue. In the absence of a well-substantiated alternative clinical diagnosis and with inconclusive serology patients were classified as indeterminate.
Categorical data are expressed as numbers and frequencies, and were compared by means of χ2 analysis. Continuous data are expressed as medians with corresponding interquartile ranges and were compared by means of the Kruskal Wallis test. Since PAI-1 plasma levels were measured using two different assays, the nonparametric regression procedure of Passing and Bablok was used to compare these two methods and to validate recalculation of one of the datasets to pool the data [14]. For this purpose, 39 samples were analysed with both assays. The Passing and Bablok regression equation resulting from this comparison is given in the Results section, together with the 95% confidence intervals for the estimates of slope and intercept. PAI-1 plasma levels obtained from the 1996–1997 cohort were converted using the regression equation. We subsequently divided PAI-1 plasma levels on admission and peak PAI-1 plasma levels during admission into tertiles of similar size. Odds ratios and the corresponding 95% confidence intervals (95% CI) were estimated by cross-tabulation using the lowest tertile as reference category. To determine the association between PAI-1 concentration and mortality independent of age, sex, year project and plasminogen activator inhhibitor-1 polymorphism, we used logistic regression analyses. A P-value ≤ 0.05 was considered to indicate statistical significance. Analyses were performed using SPSS 11.0.1. Method comparison was performed using Analyse-it Clinical Laboratory statistics module version 1.62 for Microsoft Excel.
Results
Of 233 enrolled patients with suspected dengue haemorrhagic fever or dengue shock syndrome admitted to the paediatric intensive care unit and the paediatric ward of the Dr. Kariadi Hospital, 202 (87%) were confirmed to have acute dengue, 3 (1%) were categorised as definitely not dengue and 28 (12%) as indeterminate. When a formal classification using the criteria set by the WHO was performed, 106 (52%) were classified as having dengue shock syndrome, 76 (38%) as having dengue haemorrhagic fever and 20 (10%) as having dengue fever. Nineteen of 202 patients with confirmed dengue (9%) died during follow up.
Genomic DNA was obtained from 194 patients. The remaining patients could not be typed because of insufficient volume of blood left or low yield of DNA. Clinical features and basic laboratory investigations on admission of the 194 patients are summarised in Table 1. The numbers of 4G/4G, 4G/5G and 5G/5G PAI-1 genotypes among patients in relation to mortality are summarised in Table 2. Of 192 control samples tested, 45 patients (23%) were 4G/4G homozygous, 83 (43%) were 4G/5G heterozygous, and 64 (33%) were 5G/5G homozygous. The frequencies of the PAI-1 promoter genotypes 4G/4G, 4G/5G, and 5G/5G did not differ significantly between the 1996–1997 project, the 2001–2003 project and the control group (P = 0.520). The proportion of deaths among patients with the 4G/4G, 4G/5G and 5G/5G genotype did not differ significantly (1996–1997 project: P = 0.979; 2001–2003 project: P = 0.986; two projects combined: P = 0.781). The genotype frequencies among patients with respect to final clinical diagnosis according to the criteria set by the WHO are summarised in Table 3. Results indicate that the PAI-1 promoter genotypes are not associated with dengue disease severity (P = 0.508).
Table 1 Clinical characteristics and laboratory findings
Characteristic
Age, median (IQR), years 6 (4–10)
Male sex, n (%) 92 (47)
Duration of symptoms, median (range), days 4 (1–7)
Systolic blood pressure, median (IQR), mmHg 90 (80–100)
Pulse pressure <20 mmHg, n (%) 22 (11)
Spontaneous bleeding, n (%) 118 (62)
Haematocrit, median (IQR), % 41 (36–45)
Platelet count, median (IQR), cells *103/mm3 58 (37–85)
° IQR denotes 25th and 75th interquartile range, n denotes number of patients.
Table 2 Clinical outcome of dengue virus infected patients classified by PAI-1 genotype
1996–1997 2001–2003
Genotype All patients Survivors Non-survivors All patients Survivors Non-survivors
G4/G4 8 (18%) 6 (14%) 2 (5%) 33 (22%) 32 (21%) 1 (1%)
G4/G5 15 (34%) 11 (25%) 4 (9%) 62 (41%) 60 (40%) 2 (1%)
G5/G5 21 (48%) 15 (34%) 6 (14%) 55 (37%) 53 (35%) 2 (1%)
Total 44 (100%) 32 (73%) 12 (27%) 150 (100%) 145 (97%) 5 (3%)
Table 3 Dengue disease severity and PAI-1 genotype
Genotype DF DHF DSS
G4/G4 6 (3%) 13 (7%) 22 (11%)
G4/G5 6 (3%) 35 (18%) 36 (19%)
G5/G5 8 (4%) 27 (14%) 41 (21%)
Total 20 (10%) 75 (39%) 99 (51%)
PAI-1 plasma levels were measured in 124 patients from whom sufficient frozen plasma samples was available with the use of two separate ELISA's. Thirty-nine random plasma samples were used for the comparison of the two assays. PAI-1 level measurements are known to be dependent on the assay employed, since not all antibodies used have the same specificity with respect to the various molecular forms of PAI-1 in blood (active PAI-1, PAI-1 complexed with vitronectin, inactive or latent PAI-1 and PAI-1 complexed with its target proteases tissue-type and urokinase-type plasminogen activator) [15]. Passing & Bablok regression analysis (Figure 1), however, clearly showed that the two assays employed in this study have a straightforward linear relationship between their outcomes. This allowed us to convert the PAI-1 levels measured in the 1996–1997 cohort to levels that would have been obtained with the assay used in the second cohort, using the Passing and Bablok regression equation (an intercept of -7.42 ng/ml (95% CI, -16.39 to -3.24 ng/ml) and slope of 0.35 (95% CI, 0.30 to 0.42)). Concentrations of PAI-1 on admission (71 ng/ml [42–118]) and peak values during admission (96 ng/ml [64–199]) were higher than the concentrations measured in the healthy control group (30–60 ng/ml). Patients included in the 2001–2003 project and diagnosed with DSS had higher PAI-1 plasma levels on admission (P = 0.002) and during admission (P < 0.001) than those diagnosed with DF or DHF (Table 4). However, no statistical significant difference was present between patients with different dengue disease severities when both projects were combined (PAI-1 plasma levels on admission: P = 0.212 and during admission: P = 0.089). Survival was significantly worse in patients with PAI-1 concentrations in the highest tertile (Table 5). As shown by multiple logistic regression analysis, the odds ratio for the risk of death – adjusted for age, sex, year project and plasminogen activator inhibitor-1 genotype – of PAI-1 levels at admission in the highest tertile was 7.57 (95%CI, 1.16–49.30). The adjusted odds ratio for the risk of death of peak PAI-1 levels during admission in the highest tertile was 7.41 (95%CI, 1.30–42.26). Figure 2 illustrates the relation between PAI-1 plasma concentrations and the PAI-1 genotypes. Differences were not statistically significant (values at admission: P = 0.919; peak values during admission: P = 0.470).
Figure 1 Comparison of two PAI-1 assays. Comparison of PAI-1 plasma levels obtained by a Biopool assay and an assay previously described by de Boer and colleagues [10]). The Passing & Bablok regression equation is: y = 0,3505x - 7,4227 ng/ml; n = 39. Solid line, regression line; dashed lines, 95% CI for the regression line.
Figure 2 Relation between PAI-1 plasma concentrations and PAI-1 genotype in dengue virus infected patients. Box-and-whisker plots of PAI-1 plasma concentrations on admission (A.) and peak values during admission (B.). The central line in the box plot represents the median, the boxed areas represent the interquartile ranges. Whiskers at the ends of the box show the distance from the end of the box to the largest and smallest observed values that are less than 1.5 box lengths from either end of the box.
Table 4 Relation between PAI-1 plasma levels and disease severity°
Project PAI-1 DF DHF DSS
(ng/ml) Median (IQR) Median (IQR) Median (IQR)
1996–1997¶ Admission - - 56.7 (33.0–130.2)
Peak values - - 91.6 (37.7–343.6)
2001–2003 Admission 77.0 (69.0–99.0) 60.0 (42.3–82.5) 107.0 (52.0–670.0)
Peak values 82.0 (69.0–129.0) 78 (59.3–118.3) 264.0 (117.0–844.0)
Total Admission 77.0 (69.0–99.0) 60.0 (42.3–82.5) 87.0 (35.0–180.8)
Peak values 82.0 (69.0–129.0) 78 (59.3–118.3) 130.7 (53.7–531.1)
° IQR denotes 25th and 75th interquartile range.
¶ Only DSS patients were included in the 1996 project.
Table 5 Mortality risk for plasminogen activator-1 plasma levels.*
PAI-1 ng/ml Number of patients Number of deaths OR (95% CI) ORadj (95% CI)°
Value at admission <50 41 2 1 1
50–94 40 3 1.6 (0.3–10.0) 3.3 (0.4–25.0)
>94 41 8 4.7 (0.9–23.8) 7.6 (1.2–49.3)
Maximum value during admission <72 41 2 1 1
72–151 42 3 1.5 (0.2–9.5) 2.1 (0.2–18.2)
>151 41 10 6.3 (1.3–30.8) 7.4 (1.3–42.3)
* CI, denotes confidence interval; OR, denotes odds ratio; ORadj, denotes adjusted odds ratio
° Risk of death adjusted for age, sex, year project and plasminogen activator inhhibitor-1 polymorphism. Lowest tertile was used as reference category.
Discussion
This study was undertaken to investigate the relationship between PAI-1 plasma concentrations and clinical outcome of dengue virus infections, and to establish whether PAI-1 plasma concentrations in dengue virus infected individuals are associated with the 4G/5G promotor polymorphism in the PAI-1 gene. We and others have previously found increased PAI-1 plasma concentrations in patients with severe dengue in particular in those with a poor clinical outcome [5,7]. Since a genetic predisposition to produce high PAI-1 plasma concentrations appears to be associated with poor clinical outcome in Neisseria meningitides infections [8,9], we hypothesised that dengue virus infected individuals carrying the 4G/4G genotype have higher PAI-1 plasma concentrations and are therefore at increased risk of death. Consistent with previous studies, we found increased PAI-1 concentrations in dengue virus infected individuals. However, PAI-1 plasma concentrations were not related to dengue disease severity, but were significantly associated with death from dengue. No significant association between PAI-1 plasma concentrations and carriage of the 4G/4G genotype was observed. The frequencies of the three genotypes between survivors and non-survivors, and between patients with different disease severities were not different. These findings suggest that increased PAI-1 plasma concentrations, and dengue disease severity and mortality are not dependent on the 4G polymorphism in the PAI-1 gene in this population.
An increased risk of death in dengue virus infected patients with high PAI-1 plasma concentrations adds to findings of PAI-1 levels being able to predict lethality in patients with bacterial sepsis in a very sensitive way [16-20]. One of the primary roles of PAI-1 in vivo is to inhibit tissue-type plasminogen activator, the major proteolytic activator of plasminogen [21,22]. By inhibiting fibrinolytic activity, increased PAI-1 concentrations may contribute to a procoagulant state leading to an increased deposition of fibrin and formation of microthrombi with subsequent multiorgan failure and death. A variety of cells, including endothelial cells, hepatocytes and platelets, synthesize and secrete PAI-1 in response to inflammatory stimuli such as interleukin-1 and tumour necrosis factor [10,22-24]. The release of these inflammatory mediators by monocytes and T lymphocytes activated by dengue virus may well contribute to the over-production of this inhibitor of fibrinolysis [25-27]. Dawson and colleagues showed that the common insertion/deletion (4G/5G) polymorphism in the promotor region of the PAI-1 gene affects the response of the gene to acute phase stimuli [23]. The 4G allele produced six times more mRNA than the 5G allele in response to interleukin-1 [23]. Eriksson and colleagues, however, were unable to reproduce these findings and based on their study results they concluded that the insertion/deletion (4G/5G) polymorphism is not related to an allele-specific response to interleukin-1 [28]. Instead, they found that the insertion/deletion (4G/5G) polymorphism influences basal PAI-1 transcription only [28].
Apparently other underlying mechanisms not related to the 4G/5G polymorphism must be involved in the increase in PAI-1 levels found in dengue virus infected individuals. This might include clearance impairment rather than or in addition to stimulation of synthesis. It is interesting to note that PAI-1 is cleared from the circulation by the liver [29]. Indeed in patients with severe liver disease, PAI-1 has been shown to be increased as a result of a decrease in hepatic clearance [30,31]. Since hepatic dysfunction is a relatively common finding in severe dengue virus infections, it is possible that a less efficient clearance contributes to increased PAI-1 levels in dengue virus infected individuals [32-34]. Previous studies investigated factors that could potentially influence PAI-1 levels, including environmental factors, metabolic determinants, ethnicity and a variety of other polymorphisms within the PAI-1 gene [35-38]. It remains unclear whether and to what extent these factors contribute to the variability in PAI-1 levels in dengue virus infected individuals. Previously studied individuals were either healthy, were patients with coronary artery disease, or were patients with diabetes mellitus. Clearly these study populations cannot be compared to patients suffering from a severe infectious disease that is characterised by an overwhelming inflammatory response.
Several potential limitations of the present study should be noted. The 1996–1997 cohort was characterized with a high mortality rate of 27%. Although the exact reason for this high mortality remains to be determined, it is likely that it results from a combination of factors. Our study was performed in a Tertiary Hospital that serves a large part of Middle-Java. Patients may travel long distances to be treated in this hospital and it could well be that they are presented late in the course of disease. Initial fluid resuscitation according to WHO guidelines is generally insufficient in these cases and patients usually end up in profound shock. Despite admission at the Pediatric Intensive Care Unit mortality rate is high. In addition, the 1996–1997 rainy season was characterised by high numbers of patients admitted to hospitals because of DHF/DSS and high number of non-survivors. It is believed that an unusually virulent virus circulated that year although microbiological sampling could not be performed at that time because of limited resources.
Study size is an important issue in the establishment of an association between the insertion/deletion (4G/5G) polymorphism in the PAI-1 gene and clinical outcome. Hermans and colleagues previously found an association between the homozygous 4G deletion polymorphism and mortality from Neisseria meningitides infection among 129 patients from two different cohort groups [8]. This association was also observed when the largest of the two cohort groups was studied separately, but was not seen in the smallest group in which only 37 patients were included. In order to obtain a sufficient number of patients, we therefore decided to combine the results of two different projects. This decision was based on the fact that these two projects included patients with the same ethnic background, used similar trial procedures and applied uniform diagnostic and clinical management procedures. Although mortality rates between the 1996–1997 cohort and the 2001–2003 cohort differed considerably, one must realise that in the 1996–1997 cohort only DSS patients were included. In the 2001–2003 cohort also included patients who had no evidence of circulatory failure were included. Mortality rate among DSS patients included in the 2001–2003 cohort was 18%. Our findings of similar frequencies of the PAI-1 genotypes within the 1996–1997 project and the 2001–2003 project supports our decision to combine both groups.
In conclusion, this study demonstrates that high PAI-1 plasma levels are associated with an increased risk of death from dengue without the 4G/5G polymorphism in the promotor of the gene for PAI-1 playing a role. Additional studies are needed to explore the possibility of other polymorphisms within the PAI-1 gene and factors, like ethnicity or environmental factors, contributing to the variability of PAI-1 plasma concentrations in patients with dengue.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
A.M., T.S. and P.K. wrote the first draft of the study protocol. H.C., D.B., A.O. and E.G. contributed to the writing of the study protocol. A.M. and T.S. were responsible for implementation of the study. T.S. was responsible for management of patients, and data collection at the study site. C.H., A.L., S.F. and P.R. were responsible for the measurement of PAI-1 plasma levels and for typing of the polymorphism. P.K. and A.O. were responsible for all diagnostic procedures. A.M. and A.L. did all statistical analyses. A.M., A.L., H.C., and E.G. wrote the first draft of the report, and all other authors contributed at subsequent stages.
Acknowledgements
This project was supported by a grant from the Royal Netherlands Academy of Arts and Sciences.
H. ten Cate is a Clinical Established Investigator of the Netherlands Heart Foundation.
Besides the authors, the following investigators participated in this study: Indonesia: C. Suharti, R. Djokomoeljanto (Department of Internal Medicine, dr. Kariadi Hospital, University of Diponegoro, Semarang); A. Soemantri (Department of Paediatrics, dr. Kariadi Hospital, University of Diponegoro, Semarang); K. Djamiatun (Medical Faculty, University of Diponegoro, Semarang, Indonesia). The Netherlands: J.W.M. van der Meer, W.M.V. Dolmans (Department of Internal Medicine, University Medical Centre St. Radboud, Radboud University Nijmegen, Nijmegen); Y.T. van der Heide (Clinical Chemistry and Hematology Laboratory, Slotervaart Hospital, Amsterdam); K. Stittelaar (Institute of Virology, Erasmus Medical Centre, Rotterdam); E. Vogels (Laboratory for Experimental Medicine, Academic Medical Centre, University of Amsterdam, Amsterdam); K. Joop (Hematology and Clinical Chemistry Laboratory, Onze Lieve Vrouwe Gasthuis, Amsterdam).
==== Refs
Rigau-Perez JG Clark GG Gubler DJ Reiter P Sanders EJ Vorndam AV Dengue and dengue haemorrhagic fever Lancet 1998 352 971 977 9752834 10.1016/S0140-6736(97)12483-7
Dengue Haemorrhagic Fever: Diagnosis, treatment, prevention and control 1997 2 World Health Organisation
Hales S de Wet N Maindonald J Woodward A Potential effect of population and climate changes on global distribution of dengue fever: an empirical model Lancet 2002 360 830 834 12243917 10.1016/S0140-6736(02)09964-6
Mairuhu AT Mac Gillavry MR Setiati TE Soemantri A ten Cate H Brandjes DP van Gorp EC Is clinical outcome of dengue-virus infections influenced by coagulation and fibrinolysis? A critical review of the evidence Lancet Infect Dis 2003 3 33 41 12505032 10.1016/S1473-3099(03)00487-0
van Gorp EC Minnema MC Suharti C Mairuhu AT Brandjes DP ten Cate H Hack CE Meijers JC Activation of coagulation factor XI, without detectable contact activation in dengue haemorrhagic fever Br J Haematol 2001 113 94 99 11328287 10.1046/j.1365-2141.2001.02710.x
van Gorp EC Setiati TE Mairuhu AT Suharti C Cate HH Dolmans WM van der Meer JW Hack CE Brandjes DP Impaired fibrinolysis in the pathogenesis of dengue hemorrhagic fever J Med Virol 2002 67 549 554 12116003 10.1002/jmv.10137
Wills BA Oragui EE Stephens AC Daramola OA Dung NM Loan HT Chau NV Chambers M Stepniewska K Farrar JJ Levin M Coagulation Abnormalities in Dengue Hemorrhagic Fever: Serial Investigations in 167 Vietnamese Children with Dengue Shock Syndrome Clin Infect Dis 2002 35 277 285 12115093 10.1086/341410
Hermans PW Hibberd ML Booy R Daramola O Hazelzet JA de Groot R Levin M 4G/5G promoter polymorphism in the plasminogen-activator-inhibitor-1 gene and outcome of meningococcal disease. Meningococcal Research Group Lancet 1999 354 556 560 10470700 10.1016/S0140-6736(99)02220-5
Westendorp RG Hottenga JJ Slagboom PE Variation in plasminogen-activator-inhibitor-1 gene and risk of meningococcal septic shock Lancet 1999 354 561 563 10470701 10.1016/S0140-6736(98)09376-3
de Boer JP Abbink JJ Brouwer MC Meijer C Roem D Voorn GP Lambers JW van Mourik JA Hack CE PAI-1 synthesis in the human hepatoma cell line HepG2 is increased by cytokines--evidence that the liver contributes to acute phase behaviour of PAI-1 Thromb Haemost 1991 65 181 185 1711245
Mansfield MW Stickland MH Grant PJ Plasminogen activator inhibitor-1 (PAI-1) promoter polymorphism and coronary artery disease in non-insulin-dependent diabetes Thromb Haemost 1995 74 1032 1034 8560407
Groen J Koraka P Velzing J Copra C Osterhaus AD Evaluation of six immunoassays for detection of dengue virus-specific immunoglobulin M and G antibodies Clin Diagn Lab Immunol 2000 7 867 871 11063489 10.1128/CDLI.7.6.867-871.2000
Koraka P Burghoorn-Maas CP Falconar A Setiati TE Djamiatun K Groen J Osterhaus AD Detection of immune-complex-dissociated nonstructural-1 antigen in patients with acute dengue virus infections J Clin Microbiol 2003 41 4154 4159 12958240 10.1128/JCM.41.9.4154-4159.2003
Passing H Bablok W Comparison of several regression procedures for method comparison studies and determination of sample sizes. Application of linear regression procedures for method comparison studies in Clinical Chemistry, Part II J Clin Chem Clin Biochem 1984 22 431 445 6481307
Meijer P Pollet DE Wauters J Kluft C Specificity of antigen assays of plasminogen activator inhibitor in plasma: Innotest PAI-1 immunoassay evaluated Clin Chem 1994 40 110 115 8287517
Brandtzaeg P Joo GB Brusletto B Kierulf P Plasminogen activator inhibitor 1 and 2, alpha-2-antiplasmin, plasminogen, and endotoxin levels in systemic meningococcal disease Thromb Res 1990 57 271 278 2315889 10.1016/0049-3848(90)90326-8
Kornelisse RF Hazelzet JA Savelkoul HF Hop WC Suur MH Borsboom AN Risseeuw-Appel IM van V de Groot R The relationship between plasminogen activator inhibitor-1 and proinflammatory and counterinflammatory mediators in children with meningococcal septic shock J Infect Dis 1996 173 1148 1156 8627066
Mesters RM Florke N Ostermann H Kienast J Increase of plasminogen activator inhibitor levels predicts outcome of leukocytopenic patients with sepsis Thromb Haemost 1996 75 902 907 8822584
Paramo JA Perez JL Serrano M Rocha E Types 1 and 2 plasminogen activator inhibitor and tumor necrosis factor alpha in patients with sepsis Thromb Haemost 1990 64 3 6 2274926
Pralong G Calandra T Glauser MP Schellekens J Verhoef J Bachmann F Kruithof EK Plasminogen activator inhibitor 1: a new prognostic marker in septic shock Thromb Haemost 1989 61 459 462 2678584
Kruithof EK Tran-Thang C Ransijn A Bachmann F Demonstration of a fast-acting inhibitor of plasminogen activators in human plasma Blood 1984 64 907 913 6434006
Pannekoek H Veerman H Lambers H Diergaarde P Verweij CL van Zonneveld AJ van Mourik JA Endothelial plasminogen activator inhibitor (PAI): a new member of the Serpin gene family EMBO J 1986 5 2539 2544 2430793
Dawson SJ Wiman B Hamsten A Green F Humphries S Henney AM The two allele sequences of a common polymorphism in the promoter of the plasminogen activator inhibitor-1 (PAI-1) gene respond differently to interleukin-1 in HepG2 cells J Biol Chem 1993 268 10739 10745 8388372
Ryan MP Kutz SM Higgins PJ Complex regulation of plasminogen activator inhibitor type-1 (PAI-1) gene expression by serum and substrate adhesion Biochem J 1996 314 ( Pt 3) 1041 1046 8615756
Bethell DB Flobbe K Cao XT Day NP Pham TP Buurman WA Cardosa MJ White NJ Kwiatkowski D Pathophysiologic and prognostic role of cytokines in dengue hemorrhagic fever J Infect Dis 1998 177 778 782 9498463
Green S Vaughn DW Kalayanarooj S Nimmannitya S Suntayakorn S Nisalak A Lew R Innis BL Kurane I Rothman AL Ennis FA Early immune activation in acute dengue illness is related to development of plasma leakage and disease severity J Infect Dis 1999 179 755 762 10068569 10.1086/314680
Nguyen TH Lei HY Nguyen TL Lin YS Huang KJ Le BL Lin CF Yeh TM Do QH Vu TQ Chen LC Huang JH Lam TM Liu CC Halstead SB Dengue hemorrhagic fever in infants: a study of clinical and cytokine profiles J Infect Dis 2004 189 221 232 14722886 10.1086/380762
Eriksson P Kallin B 't Hooft FM Bavenholm P Hamsten A Allele-specific increase in basal transcription of the plasminogen-activator inhibitor 1 gene is associated with myocardial infarction Proc Natl Acad Sci U S A 1995 92 1851 1855 7892190
Wing LR Hawksworth GM Bennett B Booth NA Clearance of t-PA, PAI-1, and t-PA-PAI-1 complex in an isolated perfused rat liver system J Lab Clin Med 1991 117 109 114 1899689
Hayashi T Kamogawa A Ro S Yamaguchi K Kobayashi Y Takahashi Y Murayama M Plasma from patients with cirrhosis increases tissue plasminogen activator release from vascular endothelial cells in vitro Liver 1998 18 186 190 9716229
Tran-Thang C Fasel-Felley J Pralong G Hofstetter JR Bachmann F Kruithof EK Plasminogen activators and plasminogen activator inhibitors in liver deficiencies caused by chronic alcoholism or infectious hepatitis Thromb Haemost 1989 62 651 653 2510345
Mohan B Patwari AK Anand VK Hepatic dysfunction in childhood dengue infection J Trop Pediatr 2000 46 40 43 10730040 10.1093/tropej/46.1.40
Pancharoen C Rungsarannont A Thisyakorn U Hepatic dysfunction in dengue patients with various severity J Med Assoc Thai 2002 85 Suppl 1 S298 S301 12188427
Souza LJ Alves JG Nogueira RM Gicovate NC Bastos DA Siqueira EW Souto Filho JT Cezario TA Soares CE Carneiro RC Aminotransferase changes and acute hepatitis in patients with dengue fever: analysis of 1,585 cases Braz J Infect Dis 2004 8 156 163 15361994
Festa A D'Agostino RJ Rich SS Jenny NS Tracy RP Haffner SM Promoter (4G/5G) plasminogen activator inhibitor-1 genotype and plasminogen activator inhibitor-1 levels in blacks, Hispanics, and non-Hispanic whites: the Insulin Resistance Atherosclerosis Study Circulation 2003 107 2422 2427 12719278 10.1161/01.CIR.0000066908.82782.3A
Henry M Chomiki N Scarabin PY Alessi MC Peiretti F Arveiler D Ferrieres J Evans A Amouyel P Poirier O Cambien F Juhan-Vague I Five frequent polymorphisms of the PAI-1 gene: lack of association between genotypes, PAI activity, and triglyceride levels in a healthy population Arterioscler Thromb Vasc Biol 1997 17 851 858 9157947
Henry M Tregouet DA Alessi MC Aillaud MF Visvikis S Siest G Tiret L Juhan-Vague I Metabolic determinants are much more important than genetic polymorphisms in determining the PAI-1 activity and antigen plasma concentrations: a family study with part of the Stanislas Cohort Arterioscler Thromb Vasc Biol 1998 18 84 91 9445260
Mansfield MW Stickland MH Grant PJ Environmental and genetic factors in relation to elevated circulating levels of plasminogen activator inhibitor-1 in Caucasian patients with non-insulin-dependent diabetes mellitus Thromb Haemost 1995 74 842 847 8571308
|
16274483
|
PMC1308869
|
CC BY
|
2021-01-04 16:36:23
|
no
|
Thromb J. 2005 Nov 7; 3:17
|
utf-8
|
Thromb J
| 2,005 |
10.1186/1477-9560-3-17
|
oa_comm
|
==== Front
J Transl MedJournal of Translational Medicine1479-5876BioMed Central London 1479-5876-3-411628198110.1186/1479-5876-3-41ResearchCD8+, HLA-unrestricted, cytotoxic T-lymphocyte line against malignant melanoma Somasundaram Rajasekharan [email protected] Laura [email protected] DuPont [email protected] Dorothee [email protected] The Wistar Institute, 3601 Spruce Street, Philadelphia, PA 19104, USA2 Hematology-Oncology Division, Department of Medicine, at the Hospital of the University of Pennsylvania, 34th and Spruce Streets, Philadelphia, PA 19104, USA2005 10 11 2005 3 41 41 27 10 2005 10 11 2005 Copyright © 2005 Somasundaram et al; licensee BioMed Central Ltd.2005Somasundaram et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
A CD8+ cytotoxic T lymphocyte (CTL) line was derived from the peripheral blood mononuclear cells of a patient with primary melanoma. The CD8+ CTL line specifically lysed the autologous primary melanoma cells and not the natural killer cell-sensitive K562 cells or lymphokine activated killer cell-sensitive DAUDI cells. When a large panel of human leukocyte antigen (HLA)-matched and -unmatched allogeneic melanoma, glioma, breast and colorectal carcinoma cells was tested as targets in cytolysis assays, 4 HLA-matched and two HLA-unmatched allogeneic metastatic melanoma lines were lysed by the CD8+ CTL. Lysis of autologous and allogeneic melanoma cells was dependent on the effector-to-target cell ratio. Lysis of autologous melanoma cells was not blocked by anti-HLA class I or class II antibodies, confirming that the cytolytic activity of the CD8+ CTL was HLA-unrestricted. CTL lysis of autologous melanoma cells was CD3 (T cell receptor) dependent and FAS-FAS-L, and CD1 independent. Identification of the melanoma-associated antigen recognized by the HLA-unrestricted CTL may provide a vaccine for a broad population of melanoma patients.
==== Body
Background
Cytotoxic T lymphocytes (CTL) have been established from various lymphoid sources of patients with metastatic melanoma[1]. The majority of CTL are of CD8 phenotype and generally lyse tumor cells in a human leukocyte antigen (HLA)-restricted manner. A number of melanoma-associated antigens and peptides defined by CD8+ CTL have been identified and the majority of these CTL are HLA-A2 restricted [1,2]. CD8+ CTL that recognized and lysed melanoma targets of various HLA types also have been described [3-6]. However, it is unclear whether the CTL are truly HLA-unrestricted, as the allogeneic tumor target cells used were not fully HLA subtyped [3-6] and, therefore, partial HLA matching of these targets with the CTL cannot be excluded with certainty. Furthermore, the demonstration of the absence of HLA restriction was based on the absence of CTL lysis blocking in the presence of high concentrations of anti-HLA antibodies. However, the high effector-to-target (E:T) cell ratios used in those studies, accompanied by high tumor cell lysis in control cultures [5,6] may be responsible for the absence of CTL blocking by anti-HLA antibodies.
We have established a CD8+ CTL line from the peripheral blood lymphocytes of a patient with primary melanoma. When the cytotoxic activity of the CTL line was tested against a large panel of allogeneic cell lines of melanoma, glioma, breast or colorectal carcinoma, autologous or allogeneic Epstein-Barr virus (EBV)-transformed B cells, or autologous fibroblasts, the autologous and a few allogeneic, HLA non-matched melanoma cells were lysed and induced interferon (IFN)-γ and granulocyte monocyte-colony stimulating factor (GM-CSF) secretion by the CTL. Furthermore, the lysis of the autologous or allogeneic tumor cells was not blocked by monoclonal antibodies (MAbs) to HLA-class I or class II, confirming that the CTL lyse targets in a HLA-unrestricted manner.
Methods
Patient 793
Patient 793 (male Caucasian, 39 years old) had excision of a "low risk" primary melanoma [superficial spreading type with early vertical growth phase present; the tumor thickness was 0.55 mm and the vertical growth phase had a brisk lymphoid infiltrate, with no evidence of metastases]. The primary lesion was excised ~20 years ago and there has been no recurrence since. The patient did not receive adjuvant chemotherapy after removal of the primary lesion.
Cell lines
Melanoma cell line WM793 was established from the vertical growth phase of a primary lesion of patient 793 [7]. Cell line 1205LU, the metastatic variant of WM793, was established after repeated passages of WM793 cells both in vitro and in vivo in nude mice [8]. Melanoma cell lines WM75, WM98, WM164, and WM1158 were derived from metastatic lesions of melanoma patients [7]. All cell lines were maintained in McCoy's Dulbecco 153-Leibovitz 15 (MCDB153-L15) medium (Sigma, St. Louis, MO) containing 2% fetal bovine serum (FBS). Metastatic melanoma cell line DM196 was obtained from T. L. Darrow (Duke University Medical Center, Durham, NC) and was maintained in Dulbecco's modified Eagle's medium (DMEM; GIBCO-Invitrogen, Carlsbad, CA) supplemented with 5% FBS. Metastatic melanoma cell line ME9874 was obtained from A. Anichini (Istituto Nazionale Tumori, Milan, Italy) and maintained in RPMI 1640 Glutamax medium (GIBCO-Invitrogen) supplemented with 10% FBS. Metastatic melanoma cell line A375 was obtained from American Type Culture Collection (ATCC; Rockville, MD) and maintained in RPMI 1640 medium supplemented with 10% FBS. Rectal carcinoma cell lines WC007 and WC008 were maintained in MCDB20l-L15 (Sigma) medium supplemented with 2% FBS[9]. The glioma cell lines U373MG and U87MG (obtained from Dr. Darell Bigner, Duke University Medical Center) and the breast cancer cell line MDA MB231 (obtained from Dr. Daniela Santoli, The Wistar Institute) were maintained in DMEM medium supplemented with 10% FBS. HLA class I and II expression by tumor cells was determined in complement-dependent lysis assay using HLA typing trays (One Lambda, Inc., Canoga Park, CA). 793 fibroblast cell line [10] was maintained in DMEM containing 10% FBS.
EBV-transformed B cell line was established from freshly isolated peripheral blood mononuclear cells (PBMC) of patient 793 as described before[10]. EBV-B cell lines 888, 1363, 1088, 1102, and 4226 were established from melanoma patients' PBMC. Natural Killer (NK) cell target K562 (human erythroleukemia cell line) and lymphokine-activated killer (LAK) cell target Daudi (human lymphoblastoid cell line) were obtained from ATCC. All lymphoid cell lines were maintained in RPMI 1640 medium supplemented with 10% FBS.
Antibodies
The following MAbs were used: HLA-class I-specific MAb W6/32 and HLA-class II-specific MAbs B33.1 and D1.B6 (obtained from B. Perussia, Thomas Jefferson University); MAb H24B5 to influenza virus hemagglutinin (obtained from W. Gerhard, The Wistar Institute); MAb ME36.1 to GD2 and GD3; MAb ME50.8 to GD3; MAb 31.3 to chondroitin sulfate proteoglycan; MAb MENu4B to v(e.c.)itronectin receptor; MAb 20.4 to nerve growth factor receptor; MAb 425 to epidermal growth factor receptor; MAb 451-1-C7-1 to fibroblast marker; MAbs 439-3-27-2-3 and 456-1-A11-2 to carbohydrates; MAbs ME7529 and ME77.1 to 120 kd protein on melanoma cells and MAb 507-2-2-4 to uncharacterized melanoma-associated antigen [11]. MAb HIT3a to CD3; MAb RPA-T4 to CD4; MAb RPA-T8 to CD8; MAb M-A251 to CD25; MAb 5C3 to CD40; MAb TRAP1 to CD40L; MAb L307.4 to B7-1 (CD80); MAb 2331 to B7-2 (CD86); MAb HI149 to CD1a; MAb Nok-1 to CD95L (Fas ligand), and MAb DX2 to CD95 (Fas receptor) (all antibodies were obtained from BD Pharmingen, San Diego, CA); MAb BMAO 31 to human T-cell receptor (TCR) α/β common region (T cell Sciences, Cambridge, MA); MAb OKT3 to CD3; MAbs 4F-2 and 5AG to IL-4 (DNAX, Palo Alto, CA); MAbs B133.1.1 and B133.5.1 to IFN-γ; MAbs B154.9.1 and B154.9.2 to tumor necrosis factor (TNF-α) (obtained from G. Trinchieri, The Wistar Institute); and MAb C9.1 and C16.3 to granulocyte macrophage-colony stimulation factor (GM-CSF) [12].
Generation of anti-melanoma CTL line
CTL were obtained from co-cultures of PBMC (105 cells/well of 96-well round-bottom microtiter plates) of melanoma patient 793 with irradiated (40,000 rads, Cs source) autologous melanoma cells WM793 (105 cells/well) in RPMI 1640 medium containing 10% human AB serum (Sigma), 10 mm HEPES (Sigma), L-arginine (116 mg/l; GIBCO-Invitrogen), L-asparagine (36 mg/l; GIBCO-Invitrogen), L-glutamine (216 mg/l; Invitrogen), and 2-mercaptoethanol (ME) (5 × 10-5 M; Sigma). Cultures were periodically stimulated with irradiated autologous tumor cells and irradiated autologous PBMC (days 8 and 14) or autologous EBV-B cells (from day 21 on) in RPMI 1640 medium containing partially purified IL-2 (day 3 on; 20 U/ml; Advanced Biotechnologies., Columbia, MD).
Cytotoxicity assay
This assay was performed as described previously [9,10,13]. Briefly, labeled targets [2 μCi of 51Cr (sodium chromate, specific activity 400–1200 Ci/g; NEN DuPont, Boston, MA) per 1 × 104 cells] were mixed in 96-well U-bottom microtiter plates with effector cells at various E:T ratios and incubated at 37°C for 4 hr. Supernatants were harvested and tested for 51Cr release (experimental release). For maximal release, target cells were treated with 0.3% Triton X-100 (Sigma). Spontaneous release of radioactivity by target cells was determined in the absence of effector cells. Spontaneous release of the various targets used was 10–15%. The percentage of cytotoxicity was determined by the following formula:
The targets (WC008, U373MG and U87MG) that exhibited high spontaneous 51Cr release were stained with neutral red dye and used in a modified cytotoxicity assay[10].
Blocking of cytotoxic activity
Cytotoxicity blocking assays were performed as described before [10,13]. Briefly, tumor targets were incubated with 10 μg/ml of anti-HLA-class I MAb W6/32 (IgG2a) or anti-HLA-class II MAb B33.1 and D1.B6 (IgG2a), MAb DX2 (IgG1) to Fas or HI149 (IgG1) to CD1, and CTL were incubated with 10 μg/ml of anti-CD3 MAb OKT3 (IgG2a). Isotype-matched control MAb H24B5 (IgG2a) or normal mouse IgG (predominantly IgG1) were used at 10 μg/ml. All incubations were performed for 1 hr at room temperature. Following blocking of tumor cells or CTL, cytotoxicity assays were performed as described above.
Cytokine determinations
Cultured CTL were washed twice with serum-containing RPMI 1640 medium, incubated in serum-containing medium (without stimulants or IL-2) for 10–12 hr at 37°C in a 5% CO2 incubator and washed once. CTL (104 cells/well) were stimulated with irradiated autologous or allogeneic tumor or normal cells (104–105 cells/well) in 96-well microtiter plates. Supernatants obtained from cultured CTL after 24 hr or 72 hr were tested for the presence of GM-CSF, TNF-α, IL-4, and IFN-γ.
IL-4, IFN-γ, TNF-α, and GM-CSF were measured by radioimmunoassay (RIA) as described before. [9] T cell supernatants at various dilutions were placed in antibody-coated wells (MAb 4F-2 to IL-4; MAb B133.1.1 to IFN-γ; MAb B154.9.2 to TNF-α; MAb C9.1 to GM-CSF), and binding was determined using 125I-labeled MAb specific for different determinants on the cytokines (MAb 5AG to IL-4; MAb B133.5.1 to IFN-γ; MAb B154.9.1 to TNF-α; MAb C16.3 to GM-CSF). Concentrations of IL-4, IFN-γ, TNF-α, and GM-CSF were determined using respective recombinant cytokine standard preparations (R&D Systems Inc., Minneapolis, MN).
IL-2 expression by the T cells was tested by reverse transcriptase-polymerase chain reaction (RT-PCR) as described[10].
Phenotyping of melanoma and T cells
Cultured cells were incubated with saturating concentrations (5 μg/ml) of either unlabeled or fluoresceinated or phycoerythrin(PE)-labeled MAbs detecting human lymphocyte markers (CD3, CD4, CD8, CD25, CD40, CD40L, CD80, CD86, CD95, and CD95L) or melanoma-associated antigens (ME36.1, ME50.8, 31.3, MENu4B, 20.4, 425, 451-1-C7-1, 439-3-27-2-3, 456-1-A11-2, ME7529, 77.1 and 507-2-2-4) in RPMI 1640 medium supplemented with 5% human AB serum for 45 min at 4°C. Cells were washed 3× and stained with fluorescein isothiocyanate conjugated (FITC) second antibody [goat anti-mouse F(ab)2, 1:200 dilution, ICN-Cappel, Irvine, CA]. Binding of the PE or FITC-conjugated antibodies was analyzed in the cytofluorograph (Ortho Diagnostics, Rariton, NJ). All values given in "Results" are corrected for irrelevant, isotype-matched control antibody binding.
TCR analysis by RT-PCR
TCR was analyzed as describted previously[10]. In brief, mRNA was extracted from 3 × 106 T cells using the mRNA DIRECT kit (Dynal Biotech, Lake Success, NY). RT-PCR and cDNA synthesis were performed using the SuperScript One-Step RT-PCR kit (Invitrogen, Carlsbad, CA), specific 5' primers encoding variable Vα (1–22) and Vβ (1–24) and a common C region (Cα and Cβ) 3' primer. PCR products were run on 2% agarose gels and analyzed.
Statistical analyses
Differences between experimental and control values (triplicates or quadruplicates) were analyzed for significance by Student's two-sided paired t-test.
Results
Phenotypic characteristics of T cell line
A T cell line was generated in mixed lymphocyte/tumor culture (MLTC) by stimulating the PBMC of melanoma patient 793 with irradiated autologous WM793 melanoma cells. Cultures received IL-2 from day 3 on and were restimulated with irradiated autologous tumor cells and IL-2 every week. The T cell line was >98% CD8+ (Table I). The CD8+ T cell line expressed HLA class I and II, CD3, CD25, FAS, FASL, B7-1, and B7-2, and a low percentage of the cells also expressed CD40 and CD40L (Table I). The T cell line expressed TCR α/β chains. Further, RT-PCR analysis indicated that the T cell line is oligoclonal as it expresses Vα2, Vα4, Vα14, Vα16 and Vβ6 chains. Attempts to clone CD8+ T cell line by limiting dilution were unsuccessful twice. WM793 melanoma cells expressed HLA class I and II and FAS, and a low percentage of the cells expressed FASL, B7-1, and B7-2 (Table I).
Table 1 Phenotypic markers of CD8+ CTL793 and WM793 Melanoma Cells
Percent cells positive
Parameter investigated CD8+ CTL793 WM793 melanoma cells
HLA Class I 95 97
HLA Class II 95 74
CD1 NAa 18
CD3 90 NA
CD4 0 NA
CD8 98 NA
CD25 80 0
CD95 (FAS) 98 95
CD95L (FAS-L) 89 10
CD80 (B7-1) 88 7
CD86 (B7-2) 32 3
CD40 10 3
CD40L 18 6
TCRα/β 92 NA
a NA = not applicable.
Cytolytic activity of CD8+ CTL line against various HLA-matched and non-matched tumor and non-tumor cells
The cytolytic activity of the CD8+ CTL line against autologous and allogeneic partially HLA-matched and unmatched melanoma cells was tested at various E:T ratios (6.25–50) at 8 weeks in culture. As controls, NK susceptible target (K562) and LAK sensitive target (Daudi) were used. The CTL line lysed autologous WM793 melanoma cells as well as partially HLA-matched (DM196, ME9874, A375 and WM75) and HLA-unmatched allogeneic melanoma cells (WM164 and WM1158), but not NK-sensitive K562 or LAK-sensitive Daudi cells (Fig. 1). In contrast to autlogous WM793 melanoma cells, which were non-metastatic in nude mice and lysed by CTL 793, 1205LU metastatic variant cells of WM793 cells isolated from nude mice were not lysed by the CTL (Table II). Absence of 1205LU lysis by CTL793 cannot be explained by absence of the expression of known melanoma-associated antigens or HLA class I in 1205LU cells, compared to WM793 cells (Fig. 2). However, we cannot exclude the possibility that other antigens are differentially expressed by the two melanoma cell lines. Thus, the antigen recognized by the CTL may be expressed by primary (WM793), but not metastatic variant (1205LU) melanoma cells, derived from the same patient. However, expression of HLA class II was low (<5%) in 1205LU cells compared to WM793 cells (Fig. 2). Since CTL793 are CD8+, we do not expect HLA class II to play a role in WM793 lysis by the CTL. Furthermore, there is strong evidence that CTL793 are HLA unrestricted in their capacity to lyse autologous tumor cells (see below).
Table 2 Lysis of various HLA-matched and -unmatched target cells and IFN-γ secretion by CD8+ 793CTL
Stimulator and target cell HLA subtypea % Maximal IFN-γ secretion GM-CSF secretion
Designation Origin Cyto-toxicityb (U/ml)c (U/ml)
WM793 Autologous primary melanoma A1, A29, B57[17], B35, DRB1 11, DQB1 0301 40.1d 10d 5.1d
1205LUe Autologous metastatic variant A1, A29, B57[17], B35, DRB1 11, DQB1 0301 6.6 <0.01 <0.01
DM196 Allogeneic metastatic melanoma A23(9), Aw34, B57[17], B44, DR1, DR3, DRw1, DRw52 12d 2.2d 2.5d
ME9874 Allogeneic metastatic melanoma A2, A24, B57[17], B60, Cw3, Cw7, DR7, DQ2 8.3d 5.9d 3.2d
A375 Allogeneic metastatic melanoma A1, A2, B57[17], C6 20.3d 41.2d 6.2d
WM75 Allogeneic metastatic melanoma A2, A29, B12w44, DR4, DR7 24.2d 4.6d 2.0d
WM98 Allogeneic metastatic melanoma A1, A3, B8, DR3 8.1 <0.01 <0.01
WM164 Allogeneic metastatic melanoma A24, B7, C7, DR13, DQ1, DQ6, DRw52 29.8d 1.6 2.0d
WM1158 Allogeneic metastatic melanoma A11, A24, B16, B60 (40), C3, DR13, DR4, DQ3, DQ6, DRw52, DRw53 45d <0.01 <0.01
U87MG Allogeneic glioma A2, B7, B44, Cw05, Cw07, DRB1 03, DRB115, DQB1 02, DQB1 06 0 <0.01 N.D.
U373MG Allogeneic glioma A2, B18, Cw05, DRB103 0 <0.01 N.D.
MDAMB231 Allogeneic breast carcinoma A2, B40, B41, Cw02, Cw17, DRB1 07, DRB1 13, DQB1 02, DQB1 03 0 <0.01 N.D.
WC007 Allogeneic colorectal carcinoma A1, A3, B35, DR1, DR4 0 <0.01 N.D.
WC008 Allogeneic colorectal carcinoma A1, A3, B15,B57[17], DR4 0 <0.01 N.D.
FOM708 Allogeneic melanocytes A1, A11, B7, B8, C1, C6 3.7 <0.01 <0.01
FOM723 Allogeneic melanocytes A3, A29, B7, B44 7.8 <0.01 <0.01
FOM1020 Allogeneic melanocytes A2, A3, B37, B57, C6 5.2 <0.01 <0.01
793 fibroblasts Autologous fibroblasts A1, A29, B57[17], B35, DRB1 11, DQB1 0301 N.D.f <0.01 N.D.
793 EBV-B Autologous B cells A1, A29, B57[17], B35, DRB1 11, DQB1 0301 0 1.3 N.D.
888EBV-B Allogeneic B cells A1, A24, B52, B55, C1, C7, DR15 N.D. <0.01 N.D.
1363EBV-B Allogeneic B cells A1, A2, B44, B51, C1, DR1 N.D. <0.01 N.D.
1088EBV-B Allogeneic B cells A1, A2, B8, B44, C5, DR4, DR17 N.D. <0.01 N.D.
1102EBV-B Allogeneic B cells A2, A24, B55, B62, C3, DR4, DR15 N.D. <0.01 N.D.
4226EBV-B Allogeneic B cells A24, A32, B27, B38, C3, DR4, DR15 N.D. <0.01 N.D.
K562 NK-sensitive erythroleukemia cells N.D. 0 <0.01 N.D.
Daudi lymphokine-activated killer cell-sensitive lymphoma cells N.D. 0 <0.01 N.D.
a Based on tumor cell typing, except for WM75 and WM98 (based on lymphocyte typing). HLA types of WM793 cells and those matching between WM793 and other cells are in bold.
b Lysis of labeled tumor target cells was determined in 4-hr 51Cr release assay using optimal E:T ratio of 20. Values represent means of triplicate determinations. Each cell line was tested in at least 2 different assays with similar results.
c IFN-γ and GM-CSF secretions were determined by RIA. Values represent means of triplicate determinations. Each cell line was tested in at least 2 different assays with similar results.
d p < 0.05 when compared to corresponding cpm values of spontaneous release of 51Cr by target cells and/or spontaneous release of cytokine by CTL.
e Lung metastasis obtained in nude mice by injecting autologous WM793 cells s.c.[8]
f N.D. = Not determined.
Figure 1 Cytotoxic reactivity of CD8+ CTL line against autologous melanoma cells. CD8+ CTL line was generated in MLTC by stimulating the PBMC of melanoma patient 793 with irradiated autologous WM793 melanoma cells. Cultures received IL-2 from day 3 onward, and were restimulated every 7 days for 2 months with autologous irradiated melanoma and EBV-B cells or PBMC. Cytotoxic effector cell responses were measured after 8 weeks. CTL lysis of 51Cr-labeled: A) autologous WM793 melanoma cells (○); B) allogeneic partially HLA-matched DM196 (□), ME9874 (Δ), A375 (▽), and WM75 (◇) melanoma cells; C) HLA-unmatched allogeneic melanomas WM164 (■) and WM1158 (●); D) NK cell target K562 (▲) lymphokine-activated killer cell target Daudi (◆) was determined in 4 h 51Cr-release assays using E:T ratios of 6.25–50.
Figure 2 Reactivity of anti-melanoma MAbs to melanoma cells. Melanoma cells (5 × 104) WM793 (□) and 1205LU (■) were incubated for 45 min with saturating concentrations of anti-melanoma MAbs, cells were washed to remove unbound MAbs and were further incubated with FITC conjugated anti-mouse antibody for 45 min. Cells were washed and binding of antibody was analyzed in the cytofluorograph. Results are expressed as percent positive cells.
A larger panel of allogeneic HLA-matched and -unmatched melanoma, glioma, breast and colorectal carcinoma cells were used as targets in cytotoxicity assays. Of the eight allogeneic melanoma cells used as targets in CTL assays, six (DM196, ME9874, A375, WM75, WM164, and WM1158) were significantly (at p < 0.05 level) and consistently (in 2 different assays) lysed (Fig. 1, Table II). Four of the six allogeneic melanoma cell lines that were lysed by the CTL share HLA-A1, A29 and/or B57 [17] alleles with the autologous WM793 cells. Two of the cell lines (WM164 and WM1158) lysed by the CTL (Fig. 1, Table II) were HLA-unmatched. There was no significant lysis of allogeneic partially HLA-matched melanocytes (FOM708, FOM723, FOM1020), autologous 793 EBV-B cells, allogeneic HLA-unmatched glioma cells (U87MG and U373MG), breast carcinoma (MDAMB231) or colorectal carcinoma (WC007 and WC008) cells (Table II).
Cytokine secretion of CD8+ CTL793
Supernatant from CD8+ CTL793 after 24–48 hr of stimulation with autologous melanoma cells WM793 contained significant amounts of IFN-γ, TNF-α, and GM-CSF, but not IL-2 or IL-4 (data not shown, except for IFN-γ release in Table II). IFN-γ, TNF-α, and GM-CSF were produced by the CTL, and not by irradiated autologous melanoma cells, since the supernatants derived from the melanoma cells showed no significant cytokine secretion (data not shown). To determine the specificity of cytokine release by the CTL, IFN-γ and GM-CSF production by the CTL after stimulation with various tumor and non-tumor cells was determined (Table II). All six allogeneic, partially HLA-matched melanoma cells, (DM196, ME9874, A375, WM75, WM164 and WM1158) induced significant IFN-γ and GM-CSF production by the CTL and also were lysed by the CTL (Table II). None of the cell lines derived from gliomas, breast and colorectal carcinomas, or melanocytes, fibroblasts or EBV-B lymphocytes induced IFN-γ release by the CTL (Table II).
Blocking of CTL activity of CD8+ CTL793 using various MAb
Melanoma cells WM793 express HLA-class I and II (97% and 74% of the cells positive, respectively; Table I). CTL activity against autologous WM793 cells was not blocked by pre-incubating target cells with saturating concentrations of MAb to HLA-class I (W6/32) or class II (MAb B33.1 or MAb D1.B6) (Table III), indicating that CTL activity is HLA-unrestricted. Surprisingly, these MAbs enhanced tumor cell lysis (Table III). In our related studies, these MAbs were able to block the tumor target cell lysis of other HLA-restricted CTLs [10,13]. CTL activity against partially matched allogeneic DM196 melanoma cells was not blocked by MAb to HLA class I (data not shown), indicating that CTL activity is HLA unrestricted.
Table 3 Absence of Cytotoxicity blocking of CD8+ 793 CTL line against WM793 by MAbs
MAba % WM793 target cell lysis % Inhibition of WM793 target cell lysisb
Designation Specificity Isotype
W6/32 HLA class I IgG2a 27.9 -47.6
B33.1 HLA class II IgG2a 23.8 -26.0
D1.B6 HLA class II IgG1 21.0 -11.3
DX2 FAS (CD95) IgG1 19.9 -5.3
OKT3 CD3 IgG2a 5.2 72.5c
HI149 CD1 IgG1 23.6 -28.9
H24B5d Influenza virus IgG2a 18.9 2.6
Normal mouse Igd Unknown Predominantly IgG1 18.3 3.2
a Tumor targets were incubated with MAb (saturating concentrations, 10 μg/ml) for 1 hr at room temperature and excess MAb was removed before the 4-hr 51Cr-release assay.
b Negative values indicate enhancement of lysis. Means of triplicate determinations at E:T ratios of 2.
c Value significantly (p < 0.01) different when compared to isotype matched control antibody.
d Isotype-matched control antibodies.
CTL793 express FAS-L and WM793 melanoma cells express FAS (Table I). To determine the role of FAS and FAS-L interaction in 793 CTL-mediated lysis of WM793 tumor targets, tumor cells were incubated with anti-FAS MAb DX2 before the CTL activity against the cells was tested. Anti-FAS MAb was unable to block 793 CTL lysis of WM793 target cells (Table III), indicating that CTL killing is independent of the FAS and FAS-L mediated pathway.
Low percentage of WM793 cells expresses CD1 (Table I). The role of CD1 in CTL793-mediated lysis of WM793 tumor targets was tested by incubating tumor cells with anti-CD1 MAb HI149 before the CTL activity against the cell was tested. Anti-CD1 MAb was unable to block 793 CTL lysis of WM793 target cells, indicating that CD1 is not involved in CTL-mediated lysis of tumor cells (Table III).
CTL activity of 793 CTL against WM793 cells was significantly blocked by preincubating effector cells with saturating concentrations of anti-CD3 MAb (Table III). Thus, lysis of target cells is mediated by TCR.
Discussion
There are a number of reports of HLA-unrestricted CD8+ CTL derived from the lymphocytes of melanoma patients [3-6,14]. Kubo et al [14] described two CTL clones, each expressing a single TCR (αβ) and recognizing HLA-matched tumor targets in an HLA-restricted fashion and HLA-unmatched targets in an HLA-unrestricted manner. HLA-unrestricted CTL from ovarian, breast and pancreatic carcinoma patients also have been reported and they are known to recognize epitopes of mucin (MUC1) tumor-associated antigen [15-17]. However, it is unclear whether the CTL against melanomas and carcinomas are truly HLA-unrestricted, as target cells were not fully HLA subtyped [3-6] and high E:T cell ratios were used in cytolysis assays [5,6] (see Introduction).
Our study provides strong evidence that the cytolytic activity of the CD8+ 793 CTL line is HLA-unrestricted. We have tested our CTL in two sets of experiments to confirm that the cytolytic activity is HLA-unrestricted. In the first set of experiments we have used a panel of HLA-matched and -unmatched melanoma targets in CTL lysis assays (Table II). The target cells were fully HLA subtyped. In the second set of experiments we have used anti-HLA antibodies to block the cytolytic activity of the CTL at low E:T cell ratios (Table III). CD8+ 793 CTL line lysed autologous melanoma cells, and four different partially HLA-matched (HLA-A1, A29 or HLA-B57 [17] alleles) and two unmatched allogeneic melanoma cells; none of the allogeneic tumor cell lines derived from glioma, breast or colon carcinoma were lysed. These results suggest that the CD8+ CTL lysed melanoma targets in an HLA-unrestricted manner. These results were confirmed in cytokine release assays and in assays where autologous tumor cells were incubated with MAb to HLA-class I or -class II prior to determining tumor cell lysis by the CTL at low E:T ratios. Unexpectedly, pre-incubation of tumor cells with MAb to HLA-class I or -class II resulted in enhancement of CTL tumor lysis (Table III). Similar enhancement of CTL target cell lysis in the presence of anti-HLA antibodies was observed by other investigators [3,4], although the mechanism of this phenomenon is unknown.
Our data rule out a role of Fas/Fas ligand interaction in tumor cell lysis by the CD8+ CTL line as MAb to FAS was unable to block CTL lysis of melanoma cells in spite of expression of FAS-L and FAS by the CTL and tumor cells, respectively (Table I).
The CD8+ CTL described here recognize an antigen(s) preferentially expressed by melanoma cells. This antigen was not present on allogeneic cell lines of glioma (two), breast carcinoma (one), and colorectal carcinoma (two), or allogeneic partially matched melanocytes (three) or autologous fibroblasts (one), and autologous (one) or allogeneic (five) EBV-B cells. A presumably HLA-unrestricted CTL clone recognizing a determinant shared by allogeneic melanoma and lung carcinoma cells has been described [14]. A short-term presumably HLA-unrestricted CTL line established by repeated stimulation of lymph node T cells obtained from a pancreatic carcinoma patient recognized and lysed MUC-1 positive pancreatic, breast, and ovarian carcinoma cells[15,17]. To our knowledge, MUC-1 is not expressed by melanoma cells, and thus it is highly unlikely that MUC-1 is the candidate antigen responsible for CD8+ CTL-mediated lysis.
Identification of the melanoma-associated antigen recognized by the CD8+, HLA-unrestricted CTL described here will provide a vaccine for a broader population of melanoma patients. The antigen is expressed by the majority of allogeneic, metastatic melanoma cells, although it is lost by metastatic variant cells isolated from primary, autologous WM793 melanoma cells by passaging these cells in nude mice.
List of abbreviations
CTL, cytotoxic T cell; EBV, Epstein-Barr virus; E:T, effector-to-target; FITC, fluorescein isothiocyanate; GM-CSF, granulocyte monocyte-colony stimulating factor; HLA, human leukocyte antigen; IFN-γ, interferon gamma; IL, interleukin; LAK, lymphokine-actived killer; MAb, monoclonal antibody; MLTC, mixed lymphocyte tumor culture; MUC1, mammary/pancreas mucin; PBMC, peripheral blood mononuclear cells; PE, phycoerythrin; RIA, radioimmunoassay; RT-PCR, reverse transcriptase-polymerase chain reaction; TCR, T-cell receptor; TNF, tumor necrosis factor.
Acknowledgements
We thank Dr. Giorgio Trinchieri for providing MAbs for cytokine assays, Mr. Jeffrey Faust for FACS analysis, and Mrs. Marion Sacks for editorial assistance. This work was supported by National Institutes of Health grants CA25874, CA10815 and CA93372, and the Commonwealth Universal Research Enhancement Program, Pennsylvania Department of Health.
==== Refs
Castelli C Rivoltini L Andreola G Carrabba M Renkvist N Parmiani G T-cell recognition of melanoma-associated antigens J Cell Physiol 2000 182 323 331 10653598 10.1002/(SICI)1097-4652(200003)182:3<323::AID-JCP2>3.0.CO;2-#
Renkvist N Castelli C Robbins PF Parmiani G A listing of human tumor antigens recognized by T cells Cancer Immunol Immunother 2001 50 3 15 11315507 10.1007/s002620000169
de Vries JE Spits H Cloned human cytotoxic T lymphocyte (CTL) lines reactive with autologous melanoma cells. I. In vitro generation, isolation, and analysis to phenotype and specificity J Immunol 1984 132 510 519 6197458
Fossati G Anichini A Parmiani G Melanoma cell lysis by human CTL clones: differential involvement of T3, T8 and HLA antigens Int J Cancer 1987 39 689 694 3108167
Mukherji B Guha A Chakraborty NG Sivanandham M Nashed AL Sporn JR Ergin MT Clonal analysis of cytotoxic and regulatory T cell responses against human melanoma J Exp Med 1989 169 1961 1976 2471770 10.1084/jem.169.6.1961
Topalian SL Solomon D Rosenberg SA Tumor-specific cytolysis by lymphocytes infiltrating human melanomas J Immunol 1989 142 3714 3725 2785562
Hsu MY Elder DE Herlyn M Masters JRW and Palsson B Melanoma: The Wistar Institute (WM) cell lines Human Cell Culture 1999 Great Britain, Kluwer Academic Publishers 259 274
Juhasz I Albelda SM Elder DE Murphy GF Adachi K Herlyn D Valyi-Nagy IT Herlyn M Growth and invasion of human melanomas in human skin grafted to immunodeficient mice Am J Pathol 1993 143 528 537 8342600
Jacob L Somasundaram R Smith W Monos D Basak S Marincola F Pereira S Herlyn D Cytotoxic T-cell clone against rectal carcinoma induced by stimulation of a patient's peripheral blood mononuclear cells with autologous cultured tumor cells Int J Cancer 1997 71 325 332 9139862 10.1002/(SICI)1097-0215(19970502)71:3<325::AID-IJC3>3.0.CO;2-#
Somasundaram R Robbins P Moonka D Loh E Marincola F Patel A Guerry D Herlyn D CD4+, HLA class I-restricted, cytolytic T-lymphocyte clone against primary malignant melanoma cells Int J Cancer 2000 85 253 259 10629086
Herlyn M Koprowski H Melanoma antigens: immunological and biological characterization and clinical significance Annu Rev Immunol 1988 6 283 308 3289568 10.1146/annurev.iy.06.040188.001435
Horwitz DA Gray JD Behrendsen SC Kubin M Rengaraju M Ohtsuka K Trinchieri G Decreased production of interleukin-12 and other Th1-type cytokines in patients with recent-onset systemic lupus erythematosus Arthritis Rheum 1998 41 838 844 9588735 10.1002/1529-0131(199805)41:5<838::AID-ART10>3.0.CO;2-S
Somasundaram R Jacob L Swoboda R Caputo L Song H Basak S Monos D Peritt D Marincola F Cai D Birebent B Bloome E Kim J Berencsi K Mastrangelo M Herlyn D Inhibition of cytolytic T lymphocyte proliferation by autologous CD4+/CD25+ regulatory T cells in a colorectal carcinoma patient is mediated by transforming growth factor-beta Cancer Res 2002 62 5267 5272 12234995
Kubo H Abe J Obata F Nakajima H Tsunoda M Ogawa A Nakayama S Beck Y Kohsaka T Darrow TL Abdel-Wahab Z Saida T Takiguchi M Dual recognition of a human cytotoxic T-cell clone for melanoma antigens Cancer Res 1996 56 2368 2374 8625313
Barnd DL Lan MS Metzgar RS Finn OJ Specific, major histocompatibility complex-unrestricted recognition of tumor-associated mucins by human cytotoxic T cells Proc Natl Acad Sci U S A 1989 86 7159 7163 2674949
Jerome KR Domenech N Finn OJ Tumor-specific cytotoxic T cell clones from patients with breast and pancreatic adenocarcinoma recognize EBV-immortalized B cells transfected with polymorphic epithelial mucin complementary DNA J Immunol 1993 151 1654 1662 8393050
Ioannides CG Fisk B Jerome KR Irimura T Wharton JT Finn OJ Cytotoxic T cells from ovarian malignant tumors can recognize polymorphic epithelial mucin core peptides J Immunol 1993 151 3693 3703 7690810
|
16281981
|
PMC1308870
|
CC BY
|
2021-01-04 16:39:27
|
no
|
J Transl Med. 2005 Nov 10; 3:41
|
utf-8
|
J Transl Med
| 2,005 |
10.1186/1479-5876-3-41
|
oa_comm
|
==== Front
Virol JVirology Journal1743-422XBioMed Central London 1743-422X-2-841628197210.1186/1743-422X-2-84MethodologyDevelopment of an improved microneutralization assay for respiratory syncytial virus by automated plaque counting using imaging analysis Zielinska Edyta [email protected] Daiqing [email protected] Hong-Yin [email protected] Jorge [email protected] Ruth [email protected] Da-Ping [email protected] Clinical Immunology and Virology, Wyeth Vaccines Research, Pearl River, NY. USA2005 9 11 2005 2 84 84 8 6 2005 9 11 2005 Copyright © 2005 Zielinska et al; licensee BioMed Central Ltd.2005Zielinska et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Respiratory syncytial virus (RSV) is the major cause of lower respiratory tract infection in infants and young children. Although several experimental RSV vaccines are under investigation, immuno therapy is the only treatment currently available. In assessing the immunogenicity of various vaccine formulations, a plaque reduction neutralization assay for the evaluation of RSV neutralizing antibody has been widely used. The method produces reliable results, but it is tedious and labor intensive as it relies on manual counting by laboratory personnel. To facilitate evaluation of phase II and phase III vaccine clinical trials, a more rapid, reliable and efficient neutralization assay is needed.
Results
An improved microneutralization assay for quantifying RSV neutralizing antibodies was developed using an ImmunoSpot® Series I Analyzer (Cellular Technology Ltd., Cleveland, OH) for automated plaque counting. The method is an improvement of the established classical microneutralization assay in which immunostained plaques on transparent tissue culture plates are counted manually under a dissecting microscope. Image analyzer technology allows for fully automated counting of plaques distributed throughout an entire well. Adjustments, such as the use of opaque tissue culture plates and the TMB substrate, True Blue™ (KPL, Gaithersburg, MD), were required to adapt the assay for optimal detection of plaques by the image analyzer. The suitability and the accuracy of the method for counting RSV plaques were determined by comparative testing of a reference serum and two control sera by manual and automated counting methods. The results showed that the two methods were highly correlated (R = 0.9580) and the titers generated by them were within two-fold.
Conclusion
Our results demonstrate that the semi-automated assay is rapid and reliable. It provides results within two fold to the classical plaque microneutralization assay and is readily applied to the evaluation of neutralizing antibody titers in sera obtained from epidemiology or vaccine clinical trials.
==== Body
Background
Respiratory syncytial virus (RSV) is the most common form of lower respiratory viral infection affecting infants, the elderly, and immunocompromised individuals [1]. In severe cases, it may cause complications ranging from pneumonia and bronchiolitis to death [2]. While the most severe outcomes arise in patients with weakened or underdeveloped immune systems, RSV is also gaining notoriety as an important player in annual respiratory disease epidemics among healthy adults [3]. Consequently, there is an obvious unmet need for an efficacious vaccine. The development of a vaccine will require intensive evaluation of the immune response, which can be expedited by utilizing automation.
The neutralization assay is one of the most trusted and widely used methods employed for the detection of virus-specific neutralizing antibodies [4]. The power of the neutralization assay lies in its ability to detect biologically active antibodies. While there are many methods that provide information about different aspects of the immune response (e.g. cellular immunity, genetic markers, etc.) the neutralization assay remains a proven indicator of serological immunity for many viruses. In practice, however, the plaque neutralization assay is a laborious and time-consuming procedure, making it less suitable for testing the large numbers of samples that are obtained in clinical trials. Here, we demonstrate the utility of a method that automates the most laborious and subjective part of the serum neutralization assay – the determination of plaque number. Our results show good agreement between the visual and automated high throughput counting methods for determining RSV serum neutralization antibody titers.
These results were presented previously at the VI International Symposium on Respiratory Viral Infections, March 18–21, 2004, Fort Myers, Florida [5]
Results
The data were analyzed by two general criteria: agreement and equivalence. Fig. 1. displays analysis of agreement between the two methods by plotting titers counted automatically against those counted manually. Pearson's correlation coefficient was 0.9580. Agreement between the titers obtained by the two methods was visualized by inspecting how closely the data spread around the 45-degree line (dashed line), which in this case, reflects the value of the correlation coefficient. The analysis indicated that there is a high level of agreement between the titers generated by image analysis and the standard plaque counting method.
Figure 1 Scatter plot of automatically counted titers versus manually counted titers for each sample test presented in logarithm base 4 scale. The solid line is the simple linear regression line. The dotted line indicates the 45-degree line.
By plotting the difference between titer values of each data pair against the mean of the pair, we quantified the degree of equivalence for the majority of the data (Fig. 2). In log 4 scale, a measure of 1-log is equivalent to 1 dilution or a 4-fold change. The majority of the data lay within 0.5 log (or 2-fold) of the mean (Fig. 2) indicating that the two methods show equivalence within 2-fold.
Figure 2 Difference-mean plot shows the difference of means in log 4 scale. The four populations of data grouped across the central line represent (from left to right) the low titer control serum, the reference serum tested without compliment, the reference serum tested with complement, and the high control serum group. The y-axis is represented in log4 scale the majority of the data lying within 0.5 log.
Additional analysis was performed on the largest subset of data to determine whether within assay variability would change when using the improved counting method. We evaluated all of the tests performed on the lyophilized reference serum in the presence and absence of complement and sorted the data by assay, method, and complement treatment. The range and mean of the replicate tests are depicted in Fig. 3. The difference in the mean of 136 replicates was less than 0.05 (log 4) between groups separated by complement treatment (Table 1). Table 1 indicated that the means and variability of the automatic and manual count were very similar.
Figure 3 Boxplots of RSV reference serum titers with and without complement separated by counting methods. Open "o" represents manually counted titers and "*" represents automatically counted titers. For each group of data, the line drawn through the boxes represents the group mean; the top and bottom lines of the box represent the 25th and 75th percentiles; the brackets represent the 95th percentile.
Table 1 Summary of reference serum titers in logarithm base 4 scale.
Complement Counting Method Mean Titer Standard Deviation Minimum Maximum Number
With Manual 5.02 0.215 4.65 6.01 136
Automatic 4.97 0.204 4.41 5.69 136
Without Manual 4.21 0.297 3.67 5.15 136
Automatic 4.17 0.276 3.56 5.11 136
Discussion
The data presented here demonstrate agreement and equivalence between traditional manual and automated plaque counting methods for detection of RSV neutralizing antibody titers. The 180 tests performed in the presence and absence of complement demonstrate that a wide range of titers is accurately detectable by both assays. The automated counting method does not increase the overall variability of the assay; rather variability was observed to be slightly lower with the automated method. Furthermore, we established that the two methods generated titers within 2-fold of each other. A clinically significant change of titer for an RSV patient is indicated by a 4-fold increase in titer, indicative of seroconversion [6]. Therefore, equivalence within 2-fold provides an acceptable level of confidence for the automated counting method.
The strength of this method is that it combines established plaque neutralization procedures with the technology of computerized image scanning and analysis. It has the advantage of providing more efficient and objective results while automating the most laborious and subjective aspect of the assay – plaque counting. Results can be stored as images and plaque counts indefinitely. This allows for better tracking of raw data, as is now mandated by federal and international regulatory bodies. Another important strength of this counting system is that it is capable of detecting and differentiating plaques of different morphology and thus can be used to assay many different viruses. In fact, we have already verified the capability of this system to read plaques created by mumps, influenza and other viruses (data not shown) whether in the context of determining viral potency or performing neutralization assays.
With the technological advances now available, the speed of plate scanning can be reduced further from 15 minutes/plate to approximately 2 minutes/plate. Robotic automation of plate loading can be introduced for further efficiency. The utilization of the TMB substrate staining also facilitates conservation of primary and secondary antibody stocks, which is an important factor when using viruses for which specific antibodies are not readily or commercially available. The objectivity and efficiency provided by this method of plaque counting facilitates the determination of RSV neutralizing antibody titers and can be readily applied to human epidemiology and vaccine clinical studies.
Conclusion
In this report, we describe an RSV microneutralization assay that relies on automated plaque counting and provides a more rapid and less laborious method for detecting neutralizing antibodies to RSV. It provides equivalent results to the classical plaque neutralization assay and can be used in epidemiology and vaccine clinical studies.
Materials and methods
Serum samples
Human sera provided by Intergen Bio-Diagnostics (Purchase, NY) and Bioreclamation Inc. (Hicksville, NY), were tested for anti-RSV antibody titers. Sera were selected and pooled into 3 groups according to titer. The three groups consisted of a lyophilized reference serum, prepared under the auspices of NIAID and two control sera of high and low titer. An historical in-house control standard, C587645, was also tested. The reference serum made up the majority of tests (136 tests), whereas the control sera were each tested approximately 14 times. These sources of human sera comprised the specimens evaluated in this study and were collectively tested 180 times by both methods, in the presence and absence of complement.
Virus and cells
The A2 strain of RSV was used as the challenge virus in all tests. Vero cells (ATCC Cat #CCL 34, ATCC, Rockville, MD) were cultured in EMEM with L-glutamine, 10% FBS, 1% of antibiotic/antimycotic, and non-essential amino acids. Cells were cultured on 96-well white opaque tissue culture plates (BD Falcon, Bedford, MA) for automated counting and on transparent 96-well plates (Corning-Costar, Corning, NY) for manual plaque counting 1–3 days prior to infection.
Determination of RSV antibody titers
Microneutralization
Serum samples were heat-inactivated at 56°C for 30 minutes. Four-fold serial dilutions from 1:10 to 1:10,240 were prepared in virus diluent (EMEM with L-glutamine containing 2% FBS, 2.5% HEPES (1 M) and 1% antibiotic/antimycotic, 100×). All sera were tested in the presence and absence of 10% guinea pig complement (Cambrex/BioWhittaker, Walkersville, MD) which was added to the virus diluent prior to the addition of challenge virus. Serially diluted serum was challenged with an equal volume of the RSV-A2 strain, previously titered to give 50–100 pfu per 50 μl of inoculum. The serum/virus mixtures were incubated at 37°C, 5% CO2 for 1 h.
Vero cell monolayers, prepared in 96 well plates, were infected with 50 μl/well (in duplicate) of the serum/virus mixture. Plates were centrifuged at 1 h at 2000 rpm (700 g), followed by 30 min of rocking at room temperature. Supernatants were decanted. Plates were blotted and overlaid with 0.75% methyl cellulose (4,000 cP at 2% aqueous), prepared in MEM with 2% FBS, warmed to 37°C and inoculated at 100 μl/well. Plates were incubated at 37°C, 5% CO2 for 3 days to allow for plaque formation.
Conventional staining and plaque determination for RSV neutralization
Cells infected on transparent plates were fixed with a 50%:50% methanol:ethanol mixture at room temperature for 10 min. Plates were washed with DPBS after fixing and between staining steps. Plates were incubated for 1 h at 37°C, 5% CO2 with 50 μl/well of monoclonal antibody specific for RSV-A2 F-protein (Wyeth K6-5-1) diluted to 1:1,000 in Blotto (5% non-fat milk in PBS). Peroxidase labeled secondary goat anti-mouse IgG antibody (KPL, Gaithersburg, MD) diluted 1:100 in Blotto, was incubated at 50 μl/well for 1 h at room temperature. Plaques were developed using 100 μl/well 3,3'diaminobenzidine HRP substrate (0.5 mg/ml DAB, 0.01% H2O2) prepared in DPBS and incubated at room temperature for 5 – 10 minutes. Plates were washed with tap water to stop the reaction.
Plaques were counted manually by inverting the transparent plate under a dissecting microscope. The field of the well was separated into quadrants for ease of counting. Overlapping plaques were deemed individual when lobes were apparent.
TMB staining and plaque determination for RSV neutralization
Cells infected on opaque white tissue culture plates were fixed and washed as described above. Plates were incubated for 1 h at 37°C 5% CO2 with 50 μl/well of monoclonal antibody specific for RSV-A2 F-protein (Wyeth, K6-5-1) diluted to 1:10,000 in blotto. Peroxidase labeled secondary goat anti-mouse IgG antibody (KPL, Gaithersburg, MD) diluted 1:3,000 in Blotto was added at 50 μl/well and incubated for 1 hour at room temperature. Plaques were developed using 50 μl/well of a ready to use TMB precipitate HRP substrate, True Blue™ (KPL, Gaithersburg, MD). Higher dilutions of primary and secondary antibody were used due to the increased sensitivity of the peroxidase TMB substrate. Plates were washed thoroughly with tap water to stop the reaction and dried inverted in order to minimize bleaching.
Plates were scanned and counted by the ImmunoSpot® Image analyzer from Cellular Technology Ltd. (Cleveland, OH). The software, initially designed for use in ELISPOT analysis, has been successfully employed here for plaque detection and counting. Overlapping plaques were separated using a separation tolerance parameter set by the experimenter (Fig. 4). Minimum plaque size and sensitivity to stain comprised the major parameters that could be adjusted for counting. Parameters were adjusted on each experimental day, if necessary.
Figure 4 Vero cells infected with RSV-A2 virus and stained with True Blue™ peroxidase substrate. The image shows an example of plaque differentiation by automated counting. Each "x" represents one plaque counted by the image analyzer.
Calculation of antibody titer
Titer was calculated from the average of duplicate sample wells by extrapolating the inverse dilution of serum that produced a 60% reduction of virus according to the following formula:
X = (a-b)(e-c)/(c-d) + a
where, a = log10 of dilution above the 60% reduction point, b = log10 of dilution below the 60% reduction point, c = average plaque count above the 60% reduction point (corresponds with a), d = average plaque count below the 60% reduction point (corresponds with b) and e = value of 60% reduction of average virus control count.
Statistical analysis
All titers were reported in logarithm base 4 scale in order to visually represent a difference of one dilution (of a 4-fold dilution series) as 1 log unit. Different statistical analyses were performed to assess the agreement of titers generated by two methods. In one analysis we graphically inspected the spread of the paired titers about the 45° line and computed Pearson's correlation coefficient. In another analysis, we determined the level of equivalence between the two assays, by constructing a difference-means plot [7].
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
Design and conception of the study and co-drafted the manuscript (DPY); development of the methods and co-drafted the manuscript (EZ); assisted in the development of the automated plaque counting method (DL, HYW); statistical analysis of the data (JQ); manuscript preparation and review (RR). All authors read and approved the final manuscript.
==== Refs
Collins PL Chanock RM Murphy BR Knipe DM, Howley PM Respiratory syncytial virus. In Fields Virology 2001 1 4th Philadelphia: Lippincott Williams and Wilkins 1443 1485
Han LL Alexander JP Anderson LJ Respiratory syncytial virus pneumonia among the elderly: an assessment of disease burden J Infect Dis 1999 179 25 30 9841818 10.1086/314567
Zambon M Stockton JD Clewley JP Fleming DM Contribution of influenza and respiratory syncytial virus to community cases of influenza-like illness: an observational study Lancet 2001 358 1410 1416 11705487 10.1016/S0140-6736(01)06528-X
Casals J Maramorosch K, Koprowski H Immunological techniques for animal viruses Methods in Virology 1967 III Academic Press, New York 113 194
Zielinska E Liu D Wu H Quiroz J Rappaport R Yang DP A novel high throughput microneutralization assay for respiratory syncytial virus using an ImmuneSpot Analyzer. [Abstract] VI International Symposium on Respiratory Vial Infections, March 18–21, 2004, Fort Myers, Florida
Piedra PA Jewell AM Cron SG Atmar RL Glezen WP Correlates of immunity to respiratory syncytial virus (RSV) associated-hospitalization: establishment of minimum protective threshold levels of serum neutralizing antibodies Vaccine 2003 21 3479 3482 12850364 10.1016/S0264-410X(03)00355-4
Bland JM Altman DG Statistical methods for assessing agreement between two methods of clinical measurement Lancet 307 310 1986 8 Feb 2868172
|
16281972
|
PMC1308871
|
CC BY
|
2021-01-04 16:38:57
|
no
|
Virol J. 2005 Nov 9; 2:84
|
utf-8
|
Virol J
| 2,005 |
10.1186/1743-422X-2-84
|
oa_comm
|
==== Front
World J Surg OncolWorld Journal of Surgical Oncology1477-7819BioMed Central London 1477-7819-3-731628750110.1186/1477-7819-3-73ResearchAberrant E-cadherin staining patterns in invasive mammary carcinoma Harigopal Malini [email protected] Sandra J [email protected] Melissa P [email protected] Satish K [email protected] Edi [email protected] Paul Peter [email protected] Department of Pathology and Laboratory Medicine, New York Presbyterian Hospital Weill Cornell Medical Center, New York, NY, USA2 Department of Pathology, Memorial Sloan-Kettering Cancer Center, New York, NY USA2005 14 11 2005 3 73 73 7 6 2005 14 11 2005 Copyright © 2005 Harigopal et al; licensee BioMed Central Ltd.2005Harigopal et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
E-cadherin, a cell surface protein involved in cell adhesion, is present in normal breast epithelium, benign breast lesions, and in breast carcinoma. Alterations in the gene CDH1 on chromosome 16q22 are associated with changes in E-cadherin protein expression and function. Inactivation of E-cadherin in lobular carcinomas and certain diffuse gastric carcinomas may play a role in the dispersed, discohesive "single cell" growth patterns seen in these tumors. The molecular "signature" of mammary lobular carcinomas is the loss of E-cadherin protein expression as evidenced by immunohistochemistry, whereas ductal carcinomas are typically E-cadherin positive.
Patients and methods
We report on E-cadherin immunostaining patterns in five cases of invasive mammary carcinoma
Results
These were five exceptional instances in which the E-cadherin immunophenotype did not correspond to the apparent histologic classification of the lesion. These cases which are exceedingly rare in our experience are the subject of this report.
Conclusion
Findings such as those illustrated in this study occur in virtually all biologic phenomena and they do not invalidate the very high degree of correlation between the expression of E-cadherin and the classification of breast carcinomas as ductal or lobular type on the basis of conventional histologic criteria.
==== Body
Background
The utility of the E-cadherin immunohistochemical stain to distinguish between lobular and ductal carcinomas that are difficult to classify by morphologic features alone has been well-documented in recent years [1-8]. However, greater experience in the staining patterns of lobular and ductal lesions with the E-cadherin stain has led to the discovery of rare instances of unexpected E-cadherin staining [1,6,7,9]. In this study, we report 5 cases of invasive mammary carcinoma with a striking discordance between the structural phenotype of the lesion in hematoxylin and eosin sections and immunohistochemical staining pattern for E-cadherin, a phenomenon we have termed as "aberrant".
Patients and methods
Four cases were obtained from the Pathology department of the New York Presbyterian Hospital-Weill Medical College of Cornell University and a fifth case from Memorial Sloan Kettering Cancer Center, between Jan 1999 and July 2004. All specimens were available as formalin-fixed paraffin embedded tissue blocks and hematoxylin and eosin stained slides.
For E-cadherin immunostaining, four micron-thick sections were prepared from paraffin blocks containing lesional tissue. The slides were subsequently deparaffinized in three 5-minute changes of xylene and rehydrated through graded alcohols to distilled water. Heat induced epitope retrieval was performed on paraffin sections by pretreatment in a pressure cooker using 10 mM citrate buffer pH 6.0 for two minutes. Immunohistochemical staining was performed on paraffin sections using a TechMate 500 TM automated immunostainer (Ventana Medical Systems, Inc., Tuscon, AX) according to a modified MIP protocol (Ventana Medical System, Inc.) using the ChemMate ABC peroxidase secondary detection system (Ventana Medical Systems, Inc). A monoclonal antibody to E-cadherin, clone HECD-1 (Zymed Laboratories, Inc., San Francisco, CA) was used in 1:400 dilution. The peroxidase reaction was developed using DAB chromogen provided in the kit. Sections were counterstained with hematoxylin. The hematoxylin and eosin and immunohistochemical stained slides were subsequently reviewed and histopathologic features of lesional areas recorded.
Results
A summary of the morphologic and immunohistochemical features of the following five cases is presented in Table 1.
Table 1 Histologic classification and "aberrant" E-cadherin staining patterns in invasive mammary carcinomas.
Case In-situ E-cadherin Invasive E-cadherin Metastasis E-cadherin
1 Lobular Negative Ductal Negative Lobular Negative
2 Ductal Pos, ++ Ductal Pos, + Lobular Pos, +++
Lobular Negative
3 Lobular Negative Ductal Pos, + CK-pos Not Done
Ductal Negative
4 Lobular Not Done Lobular Pos, +++* Lobular Pos, +++
5
Lobular
Negative
Lobular
Negative
None
*in pleomorphic areas. Weak, discontinuous positivity was seen in more classical areas of invasive carcinoma in the same tumor.
Pos-positive; + weak, ++ moderate, and +++ strong cell membrane immunoreactivity.
Case 1
A 62 year-old woman underwent a left lumpectomy in April of 2001 which revealed invasive carcinoma, histologically classified as ductal type.
Microscopic findings
The 2.3 cm invasive carcinoma was well to moderately differentiated with intermediate nuclear grade. Well-formed ducts comprised greater than 90% of the invasive carcinoma (Figure 1A). In some areas, tumor cells assumed a linear growth pattern and contained intracytoplasmic mucin vacuoles (Figure 1A, inset). The invasive carcinoma was estrogen and progesterone receptor positive by immunohistochemistry. Lobular carcinoma in-situ (LCIS), mostly of the classical type with some foci having pleomorphic features and apocrine cytology was also present. In areas, classical LCIS extended into adjacent ducts in a pagetoid distribution. A single axillary lymph node was involved by metastatic carcinoma composed of tumor cells in a diffuse, discohesive pattern characteristic of metastatic lobular carcinoma. Focal gland formation was also evident in this metastatic deposit. Intracytoplasmic mucin vacuoles were readily identified within tumor cells.
Figure 1 Absence of E-cadherin in invasive "ductal" carcinoma (case 1). A: Moderately differentiated invasive ductal carcinoma. Focally, tumor cells are seen growing in a linear fashion and contain intracytoplasmic mucin vacuoles (inset) (Hematoxylin and eosin). B: Invasive ductal carcinoma exhibiting complete absence of reactivity for E-cadherin. Normal duct serves as an internal positive control (Immunoperoxidase stain for E-cadherin). C: Metastatic carcinoma with glandular differentiation (Hematoxylin and eosin, left) showing complete absence of E-cadherin immunoreactivity (Immunoperoxidase stain for E-cadherin, right).
An E-cadherin immunohistochemical stain was performed on representative samples of invasive, in-situ, and metastatic carcinoma. No cell membrane reactivity was demonstrated in the primary invasive carcinoma or in the nodal metastasis including areas of glandular differentiation (Figure 1B and 1C). Complete absence of E-cadherin immunoreactivity was also seen in foci of LCIS.
In this case, aberrant E-cadherin protein expression was represented by absence of cell membrane immunoreactivity in gland-forming regions of invasive as well as metastatic carcinoma which appeared to be phenotypically "ductal".
Case 2
A 60 year-old female presented with a left breast mass. Subsequent excisional biopsy revealed invasive duct carcinoma.
Microscopic findings
The lumpectomy specimen contained an 8 mm invasive well-differentiated duct carcinoma. Although greater than 90% of the tumor exhibited glandular differentiation, very focal areas with a single cell and linear growth pattern were also present (Figure 2A). The invasive carcinoma was estrogen and progesterone receptor positive and negative for HER-2/neu protein overexpression by immunohistochemistry. In-situ carcinoma was ductal in type with cribriform architecture and low nuclear grade arising predominantly in a background of typical and atypical columnar cell hyperplasia. Classical LCIS that expanded lobules with central discohesion was also present. The re-excisional specimen revealed additional foci of LCIS and no invasive carcinoma. An ipsilateral sentinel lymph node biopsy was performed. The lymph node was involved by metastatic carcinoma with a diffuse, discohesive growth pattern characteristic of lobular carcinoma. A minority of the tumor (<5%) also showed glandular differentiation.
Figure 2 E-cadherin reactivity in metastatic "lobular" carcinoma (case 2). A: Well-differentiated invasive ductal carcinoma (Hematoxylin and eosin, left) exhibiting weak cell membrane positivity for E-cadherin (Immunoperoxidase stain for E-cadherin, right). B: Nodal metastasis demonstrating strong cell membrane immunoreactivity for E-cadherin; higher magnification of tumor cells (inset) (Immunoperoxidase stain for E-cadherin).
The immunohistochemical stain for E-cadherin showed weak cell membrane reactivity in areas of the primary invasive carcinoma exhibiting a ductal phenotype (Figure 2A). In contrast, areas of invasive carcinoma with a linear, "lobular" phenotype displayed strong, uniform cell membrane staining (not shown). Diffuse, strong and uniform immunoreactivity was also seen in the nodal metastasis (Figure 2B).
Foci of intraductal carcinoma demonstrated moderate positivity for E-cadherin while those of classical LCIS were negative for this antigen.
In this case, aberrant E-cadherin protein expression consisted of weak immunoreactivity in the "ductal" areas of invasive carcinoma, whereas strong positivity was appreciated in areas of invasive and metastatic carcinoma with a lobular histologic phenotype.
Case 3
A 62 year-old woman presented with a right breast mass. Stereotactic needle core biopsy revealed an invasive, moderately differentiated carcinoma that was classified as ductal. Subsequent wide local excision included the needle core biopsy site as well as an area of radiographic density not previously sampled.
Microscopic findings
The excisional biopsy revealed two separate invasive tumors measuring 8 and 5 mm, respectively. Well-formed glands comprised 80–90% of both tumors. Adjacent classical LCIS was also present. The larger tumor was strongly positive for estrogen and progesterone receptors by immunohistochemistry. No membrane expression for HER-2/neu was appreciated. A single ipsilateral sentinel lymph node contained several cytokeratin-positive cells. Five additional axillary lymph nodes were free of metastatic disease.
The immunohistochemical stain for E-cadherin of the larger tumor was weak and discontinuous (Figure 3A). There was no E-cadherin cell membrane reactivity in the smaller invasive carcinoma (Figure 3B) or in adjacent LCIS (not shown). An E-cadherin stain was not performed on the sentinel lymph node.
Figure 3 Weak and absent E-cadherin reactivity in two concurrent invasive "ductal" carcinomas (case 3). A: Larger invasive ductal carcinoma exhibiting weak staining for E-cadherin. Normal duct serves as an internal positive control (Immunoperoxidase stain for E-cadherin). B: Smaller invasive ductal carcinoma showing complete absence of staining for E-cadherin. Normal duct serves as an internal positive control (Immunoperoxidase stain for E-cadherin).
Aberrant E-cadherin protein expression in this case consisted of predominantly negative immunoreactivity in invasive tumors exhibiting a ductal histological phenotype.
Case 4
A 57 year-old woman presented with a mass in the left breast.
Microscopic findings
The excised specimen contained a 2.9 cm invasive and in-situ poorly differentiated carcinoma. More than 90% of the tumor consisted of invasive carcinoma which exhibited predominantly alveolar and solid growth patterns, absence of gland formation, and intermediate-high nuclear grade (Figure 4A, left). In-situ carcinoma comprised the remainder of the tumor mass and had a solid growth pattern, apocrine features, and intermediate-high nuclear grade. Both invasive and in-situ carcinoma formed signet-ring cells containing bluish-tinged, intracytoplasmic mucin vacuoles and together with the nuclear atypia, had the appearance of pleomorphic lobular carcinoma. The invasive carcinoma showed nuclear reactivity for estrogen and progesterone receptors by immunohistochemistry. The subsequent left mastectomy specimen revealed residual invasive carcinoma. A single, 2.2 cm left axillary lymph node contained metastatic carcinoma with extranodal extension. Morphologically, the metastatic deposit showed a "lobular phenotype" in the form of signet-ring cells growing in a diffuse distribution and absence of glandular differentiation (Figure 4C, left).
Figure 4 E-cadherin reactivity in invasive "lobular" carcinoma (case 4). A: Invasive carcinoma with pleomorphic lobular features showing predominantly alveolar and solid growth patterns and intermediate-high nuclear grade (Hematoxylin and eosin, left). Strong cell membrane E-cadherin reactivity in tumor cells and in normal adjacent ductal epithelium (Immunoperoxidase for E-cadherin, right). B: Some tumor cells demonstrate a punctate staining pattern within the cell membrane or in the cytoplasm (Immunoperoxidase for E-cadherin). C: Metastatic carcinoma with lobular characteristics involving an ipsilateral axillary lymph node (Hematoxylin and eosin, left). Tumor cells showing strong cell membrane reactivity for E-cadherin (Immunoperoxidase for E-cadherin, right).
An E-cadherin immunohistochemical stain performed on the primary invasive carcinoma revealed a heterogeneous staining pattern. Tumor areas with alveolar and solid growth patterns exhibited strong, continuous cell membrane immunoreactivity (Figure 4A, right) while those having a linear growth pattern had relatively weaker, discontinuous positivity (not shown). Additionally, some tumor cells demonstrated a punctate staining pattern either within the cell membrane or cytoplasm (Figure 4B). Overall, the E-cadherin reactivity was interpreted as positive in areas of invasive carcinoma despite heterogeneous staining in some areas. Metastatic carcinoma involving the lymph node demonstrated strong, continuous membrane staining (Figure 4C, right).
Aberrant staining in this case consisted of strong E-cadherin protein expression in areas of primary invasive carcinoma morphologically resembling pleomorphic lobular carcinoma. Weaker, discontinuous E-cadherin positivity was observed in other areas of the invasive carcinoma resembling classical type of lobular carcinoma. Punctate membrane or cytoplasmic staining by E-cadherin was also observed in some tumor cells. The lymph node metastasis which also had a "lobular" appearance had strong membrane E-cadherin reactivity similar to that seen in the primary invasive carcinoma.
Case- 5
A 37 year-old woman presented with a left breast mass. Stereotactic needle core biopsy revealed in-situ and invasive, moderately differentiated carcinoma that was classified as ductal. Subsequent wide local excision of the mass was performed.
Microscopic findings
The excisional biopsy revealed an invasive tumor measuring 1.9 cm. The invasive carcinoma predominantly exhibited a classical lobular pattern with well-formed glands comprising 20% of the tumor. Extensive multifocal classical LCIS, focally florid type was also present. The tumor was positive for estrogen and progesterone receptors by immunohistochemistry. No membrane expression for HER-2/neu was present.
There was no E-cadherin cell membrane reactivity in the invasive component, both phenotypically lobular and ductal areas (Figure 5) as well as in adjacent LCIS (Figure 5A).
Figure 5 Invasive lobular carcinoma with glandular "ductal" differentiation (case 5). A: Moderately differentiated invasive "ductal" carcinoma and carcinoma in situ on core biopsy (Hematoxylin and eosin, right) showing complete absence of E-cadherin immunoreactivity. (Immunoperoxidase stain for E-cadherin, left). B: No immunoreactivity for E-cadherin (Immunoperoxidase stain for E-cadherin, left) confirms invasive lobular carcinoma with glandular differentiation in the excision biopsy (Hematoxylin and eosin, right). C: Invasive lobular carcinoma with gland formation (Hematoxylin and eosin, right and immunoperoxidase stain for E-cadherin, left).
Aberrant E-cadherin protein expression in this case consisted of complete absence of immunoreactivity in an invasive tumor with areas of unquestionable glandular differentiation.
Discussion
E-cadherin is a transmembrane glycoprotein that mediates calcium- dependent cell-cell adhesion and is expressed mainly in epithelial cells and thought to play a critical role in epithelial differentiation and morphogenesis [10,11]. Mutations in the E-cadherin gene (CDHI) located on chromosome 16q22.1 have been demonstrated in gastric, ovarian, endometrial and thyroid carcinomas in addition to lobular breast carcinomas [12]. More recently molecular alteration in the E-cadherin gene resulting in loss of expression of E-cadherin in in situ and invasive lobular carcinomas has been demonstrated by molecular studies [12-16]. Somatic mutations in CDH1 occur in lobular breast carcinomas with a frequency ranging from 10–56% (15–20% of invasive lobular carcinomas) and are rare in ductal carcinomas [17,18]. Molecular alteration with loss of heterozygosity at 16q21.1 is the most frequent chromosome alteration in lobular carcinoma and correlates with the loss of E-cadherin expression [19]. In lobular carcinomas and diffuse type of gastric carcinomas absence of E-cadherin expression has been suggested to contribute to discohesive, single cell and infiltrative growth characteristic of these tumors [20,21].
Several studies have demonstrated the reliability of E-cadherin as a marker for distinguishing ductal from lobular carcinoma. In a study by Acs et al, where 183 invasive (duct, lobular and mixed) and 198 in situ carcinomas were studied, all in situ and invasive ductal carcinomas showed strong membrane E-cadherin expression. Forty-one of 42 invasive and 50 of 53 in situ lobular carcinoma showed complete absence of E-cadherin expression [1]. Moll et al, studied 89 primary infiltrating carcinomas immunohistochemically using an antibody to E-cadherin and found that 78% of well and moderately differentiated invasive duct carcinomas (IDC) showed strong linear staining for E-cadherin. Fifty-four percent of the poorly differentiated IDC however, had reduced and heterogeneous staining. E-cadherin reactivity was absent in 86.4% of the invasive lobular carcinomas and there was weak staining in 13.6 %. Among the in situ carcinomas the majority of DCIS had strong E-cadherin expression and all cases of LCIS showed absence of staining [7].
We have separately studied a series of 132 breast carcinomas, both in situ and invasive, for E-cadherin expression [22]. In this series there was 100% correlation between classification as ductal or lobular on the basis of conventional histologic criteria and E-cadherin reactivity. In the course of this study we observed some examples of in situ lobular carcinoma involving ducts and lobules which contained residual non-neoplastic E-cadherin positive epithelial and myoepithelial cells. In some instances, especially when there was pagetoid spread of LCIS, staining of non-neoplastic cells could be mistaken for reactivity in the in situ carcinoma. We also have seen cases where in situ lobular carcinoma displayed fragmented, discontinuous and usually weak reactivity. These phenomena probably account for most of the reported instances of lobular carcinoma with E-cadherin reactivity.
Few studies have specifically investigated invasive carcinomas with mixed duct and lobular features (IDLC). Acs et al, investigated 41 cases which were classified as such due to the presence of a "mixture in variable degrees of ductal cytology and growth pattern (large, pleomorphic cells with cohesive cellular arrangement with or without lumen formation) and typical lobular pattern (dispersed infiltrating fashion) in the same lesion, or to occasional tubules or small nest formation in a lesion that otherwise showed a typical lobular type of infiltration and cytology" [1]. They correlated the E-cadherin immunohistochemical staining patterns with morphologic features (presence/absence of tumor cell nests and trabeculae, tubule/lumen formation, intracytoplasmic lumina, discohesion) in these cases and found three patterns of E-cadherin expression. Ten cases grouped as "lobular like" IDLC showed similar absence of staining as seen in traditional invasive lobular carcinomas while twenty-four "ductal like" IDLCs had uniform membrane positivity similar to that seen in typical ductal carcinomas. The accompanying in situ carcinoma was mainly LCIS and DCIS, respectively, in these groups. A third group comprising the remaining 7 cases deemed as the "intermediate" IDLC, demonstrated focal, complete loss of E-cadherin staining. These tumors were accompanied by LCIS and DCIS. Statistical analysis showed that complete loss of immunostaining correlated well with the histologic impression of lobular features and lack of tubule or lumen formation exhibited by IDLC. Acs et al concluded that all 10 "lobular like" and 24 "ductal like" examples of infiltrating carcinomas demonstrated E-cadherin expression similar to typical lobular and ductal carcinomas, respectively, and therefore, could be further classified based on their immunophenotype for E-cadherin. However, 7 (3.8%) of 182 invasive carcinomas in their study remained histologically and immuno phenotypically mixed ductal and lobular carcinoma [1]. These latter cases demonstrated real heterogeneity in E-cadherin expression, however, the authors noted that areas of tumor showing complete lack of membrane staining correlated well with the histologic impression of lobular features exhibited by the tumor.
Goldstein et al, investigated 80 mixed (lobular and ductal) breast carcinomas. They found that the percentage of strongly E-cadherin-reactive lobular carcinoma cells was greatest in the mixed, predominantly ductal carcinomas and the greatest numbers of strongly E-cadherin-stained lobular carcinoma cells were identified near the periphery of the mixed carcinomas, particularly along the leading edge of these tumors. In addition, they observed that the nuclear grade of the lobular carcinoma component increased as the percentage of the carcinoma that was of the ductal type increased [9]. In the small series that we studied [22], four carcinomas with duct and lobular features were E-cadherin positive. These divergent results indicate that further work needs to be done to determine how E-cadherin immunostaining can be used for the classification of structurally heterogeneous tumors.
The cases described in the present study represent a phenomenon wherein the histological structural phenotype did not match the "genotype" as suggested by the staining pattern of the E-cadherin immunohistochemical stain. This "aberrant" staining pattern is exceedingly rare in our experience. In cases 1, 2, 3, and 5 invasive carcinomas with predominantly gland forming elements and therefore ductal phenotype lacked E-cadherin reactivity. The differential diagnostic option in cases 1–3 and 5 is tubulolobular carcinoma based on the hybrid morphology of gland (tubules) and linear (lobular) growth pattern. Tubulolobular carcinoma is a distinct subtype of mammary carcinoma, which were originally thought to represent a tubular variant of lobular carcinoma. Few studies have examined the immunoprofile of tubulolobular carcinoma with respect to the E-cadherin staining pattern. In a recent study by Wheeler et al in which 27 cases of tubulolobular carcinoma composed of intermixed, round to angulated tubules and single file cell cords with diffuse and targetoid growth pattern demonstrated strong and diffuse positivity for E-cadherin in both the cell cord and tubular components [23]. In our experience, mammary carcinomas with duct and lobular features also showed E-cadherin positivity. The diffuse E-cadherin positive staining in tubulolobular mammary carcinoma appears to support a ductal differentiation rather than lobular origin However, in cases 1–3 and 5 the E-cadherin immunostaining in the glandular components was negative or weakly positive. In the fourth case, the poorly differentiated invasive carcinoma with pleomorphic lobular features exhibited an unusual and heterogeneous pattern of E-cadherin expression. In addition to strong cell membrane positivity, there was also cytoplasmic as well as membrane dot-like E-cadherin staining of tumor cells.
Cytoplasmic E-cadherin reactivity has been described in diffuse type of gastric adenocarcinomas as well as in some invasive lobular carcinomas [13,21,24]. Acs and co-workers encountered a single case of E-cadherin-positive invasive lobular carcinoma which was histologically compatible with pleomorphic lobular carcinoma and associated with intermediate grade solid DCIS. The authors surmised that this tumor was most likely an example of ductal carcinoma with a dispersed growth pattern. In addition, peri-nuclear dot staining by E-cadherin was seen in two cases of LCIS. Such aberrant staining may be due to mutant E-cadherin, which is incorrectly processed within the Golgi apparatus, or from accelerated protein turnover [1]. Kowalski et al., reported that all eight invasive lobular carcinomas showed only cytoplasmic staining for E-cadherin which was also localized to the cytoplasm in the majority of (7/9) metastatic lobular carcinomas [21]. In gastric examples, it has been suggested that cytoplasmic localization was due to abnormal transport mechanisms which were also present in the non-malignant gastric epithelium. This observation was taken to be evidence that alteration in E-cadherin occurs early in the development of gastric carcinogenesis. This finding was not observed in the non-neoplastic breast tissue [24].
Molecular studies have shown that invasive breast cancer is a disease with multiple cytogenetic subclones which are also present in the preinvasive lesions [25]. The two major types of breast cancer exhibit genetic heterogeneity, but occasionally a small percentage of DCIS cases are accompanied by invasive lobular carcinoma and similarly, LCIS cases may develop invasive carcinoma of the ductal or lobular type [26]. Transition from LCIS towards ductal invasive carcinoma or from DCIS towards lobular invasive carcinoma is possible [25]. Molecular analyses of pure tubular carcinomas using comparative genomic hybridization suggest that lobular and tubular carcinomas share an early genomic change, with a similar mechanism of progression [27]. The aberrant staining patterns reported in this study may be due to non-functional E-cadherin protein. Disruption of E-cadherin signaling with inactivation of E-cadherin protein by phosphorylation by activation of SRC family kinases and several growth factor receptors has been previously reported. [28]. Normal E-cadherin expression can be seen despite compromised functional ability via defects in catenin (alpha, beta and gamma) [29] Loss of E-cadherin expression can occur by gene deletion, as well as defects in transcription and methylation [30]. E-cadherin gene transcription is inhibited by two zinc-finger transcription factors SLUG and SNAIL, which may cause transient down-regulation and up-regulation of E-cadherin in primary and metastatic breast carcinoma [31].
We speculate that E-cadherin may play a critical role in molecular classification of breast carcinoma and may further elucidate the origin of tumors with a mixed or discordant phenotype.
Conclusion
Findings such as those illustrated in this study occur in virtually all biologic phenomena and they do not invalidate the very high degree of correlation between the expression of E-cadherin and the classification of breast carcinomas as ductal or lobular type on the basis of conventional histologic criteria. Further studies are necessary to clarify whether E-cadherin protein is non-functional or truly represents exceptional biology in breast carcinomas exhibiting aberrant staining of E-cadherin.
Abbreviations
IDLC – invasive carcinomas with mixed duct and lobular features
LCIS – lobular carcinoma in situ
DCIS – ductal carcinoma in situ
IDC – infiltrating ductal carcinoma
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MH, SS, MM, and PPR: These authors contributed equally to the preparation of this manuscript.
ST and EB: Each of these authors contributed one case.
==== Refs
Acs G Lawton TJ Rebbeck TR LiVolsi VA Zhang PJ Differential expression of E-cadherin in lobular and ductal neoplasms of the breast and its biologic and diagnostic implications Am J Clin Pathol 2001 115 85 98 11190811 10.1309/FDHX-L92R-BATQ-2GE0
Bratthauer GL Moinfar F Stamatakos MD Mezzetti TP Shekitka KM Man YG Tavassoli FA Combined E-cadherin and high molecular weight cytokeratin immunoprofile differentiates lobular, ductal, and hybrid mammary intraepithelial neoplasia Hum Pathol 2002 33 620 627 12152161 10.1053/hupa.2002.124789
Gamallo C Palacios J Suarez A Pizarro A Navarro P Quintanilla M Cano A Correlation of E-cadherin expression with differentiation with differentiation grade and histologic type in breast carcinoma Am J Pathol 1993 142 987 993 7682767
Goldstein NS Bassi D Watts JC Layfield LJ Yaziji H Gown AM E-cadherin reactivity of 95 noninvasive ductal and lobular lesions of the breast. Implications for the interpretation of problematic lesions Am J Clin Pathol 2001 115 534 42 11293901 10.1309/B0DD-4M7H-GJG1-7KCW
Jacobs TW Pliss N Kouria G Schnitt SJ Carcinomas in situ of the breast with indeterminate features: role of E-cadherin staining in categorization Am J Surg Pathol 2001 25 229 36 11176072
Maluf HM Swanson PE Koerner FC Solid low-grade in-situ carcinoma of the breast: role of associated lesions and E-cadherin in differential diagnosis Am J Surg Pathol 2001 25 237 244 11176073
Moll R Mitze M Ferixen UH Birchmeier W Differential loss of E-cadherin expression in infiltrating ductal and lobular breast carcinoma Am J Pathol 1993 143 1731 1742 8256859
Wahed A Connelly J Reese T E-cadherin expression in pleomorphic lobular carcinoma: an aid to differentiation from ductal carcinoma Ann Diagn Pathol 2002 6 349 351 12478484 10.1053/adpa.2002.36660
Goldstein NS Does the level of E-cadherin expression correlate with the primary breast carcinoma infiltration pattern and type of systemic metastases? Am J Clin Pathol 2002 118 425 434 12219785 10.1309/JMRD-W08Y-6K8M-7AD8
Deman JJ Van Larebeke NA Bruyneel EA Bracke ME Vermeulen SJ Vennekens KM Mareel MM Removal of sialic acid from surface of human MCF-7 mammary cancer cells abolishes E-cadherin – dependent cell-cell adhesion in an aggregation assay In Vitro Cell Dev Biol Anim 1995 31 633 639 8528519
Gumbiner BM Regulation of cadherin adhesive activity J Cell Biol 2000 148 399 404 10662767 10.1083/jcb.148.3.399
Droufakou S Deshmane V Roylance R Hanby A Tomlinson I Hart IR Multiple ways of silencencing E-cadherin gene expression in lobular carcinoma of the breast Int J Cancer 2001 92 404 408 11291078 10.1002/ijc.1208
De Leeuw WJ Berx G Vos CB Peterse JL Van de Vijver MJ Litvinov S Van Roy F Cornelisse CJ Cleton-Jansen AM. Simultaneous loss of E-cadherin and catenins in invasive lobular breast cancer and lobular carcinoma in situ J Pathol 1997 183 404 411 9496256 10.1002/(SICI)1096-9896(199712)183:4<404::AID-PATH1148>3.0.CO;2-9
Huiping C Sigurgeirsdottir JR Jonasson JG Eiriksdottir G Johannsdottir JT Egilsson V Ingvarsson S Chromosome alterations and E-cadherin gene mutations in human lobular breast cancer Br J Cancer 1999 81 1103 1110 10584868 10.1038/sj.bjc.6690815
Kanai Y Oda T Tsuda H Ochiai A Hirohashi S Point mutations of the E-cadherin gene in invasive lobular carcinoma of the breast Jpn J Cancer Res 1994 85 1035 1039 7961105
Vos CB Cleton-Jansen AM Berx G de Leeuw WJ ter Haar NT van Roy F Cornelisse CJ Peterse JL van de Vijver MJ E-cadherin inactivation in lobular carcinoma in situ of the breast: an early event in tumorigenesis Br J Cancer 1997 76 1131 1133 9365159
Berx G Cleton-Jansen AM Strumane K de Leeuw WJ Nollet F van Roy F Cornelisse C E-cadherin is inactivated in a majority of invasive human lobular breast cancers by truncation mutations throughout its extracellular domain Oncogene 1996 13 1919 1925 8934538
Kashiwaba M Tamura G Suzuki Y Maesawa C Ogasawara S Sakata K Satodate R Epithelial-cadherin gene is not mutated in ductal carcinomas of the breast Jpn J Cancer Res 1995 86 1054 1059 8567396
Etzell JE Devries S Chew K Florendo C Molinaro A Ljung BM Waldman FM Loss of chromosome 16q in lobular carcinoma in-situ Hum Pathol 2001 32 292 296 11274638 10.1053/hupa.2001.22759
Becker KF Atkinson MJ Reich U Becker I Nekarda H Siewert JR Hofler H E-cadherin gene mutations provide clues to diffuse type gastric carcinomas Cancer Research 1994 54 3845 52 8033105
Kowalski PJ Rubin MA Kleer CG E-cadherin expression in primary carcinomas of the breast and its distant metastases Breast Cancer Res 2003 5 217 22 10.1186/bcr651
Harigopal M Shin SJ Rosen PP E-cadherin reliably distinguishes lobular and ductal carcinoma [abstract] Mod Pathol 2003 16 32A
Wheeler TT Tai LH Bratthauer GL Waldner DL Tavassoli FA Tubulolobular Carcinoma of the Breast, An Analysis of 27 cases of a tumor with Hybrid Morphology and Immunoprofile Am J Surg Pathol 2004 28 1587 1593 15577677
Carpenter PM Al-Kuran RA Theuer CP Paranuclear E-cadherin in gastric adenocarcinoma Am J Clin Pathol 2002 118 887 894 12472282 10.1309/EKFB-0HJT-AB1D-5LJB
Buerger H Simon R Schafer KL Diallo R Littmann R Poremba C van Diest PJ Dockhorn-Dworniczak B Bocker W Genetic relation of lobular carcinoma in-situ, ductal carcinoma in-situ, and associated invasive carcinoma of the breast Mol Pathol 2000 53 118 121 10897329 10.1136/mp.53.3.118
Maluf H Koerner F Lobular carcinoma in situ and infiltrating ductal carcinoma: frequent presence of DCIS as a precursor lesion Int J Surg Pathol 2001 9 127 131 11484500
Waldman FM Hwang ES Etzell J Eng C DeVries S Bennington J Thor A Genomic alteration in tubular breast carcinomas Hum Pathol 2001 32 222 226 11230710 10.1053/hupa.2001.21564
Behrens J Vakaet L Friis R Winterhager E Van Roy F Mareel MM Birchmeier W Loss of epithelial differentiation and gain of invasiveness correlates with tyrosine phosphorylation of the E-cadherin/beta-catenin complex in cells transformed with a temperature-sensitive v-SRC gene J Cell Biol 1993 120 757 766 8425900 10.1083/jcb.120.3.757
Shimoyama Y Nagafuchi A Fujita S Gotoh M Takeichi M Tsukita S Hirohashi S Cadherin dysfunction in a human cancer cell line: possible involvement of loss of alpha-catenin expression in reduced cell-cell adhesiveness Cancer Res 1992 52 5770 5774 1394201
Ji X Woodard AS Rimm DL Fearon ER Transcriptional defects underlie loss of E-cadherin expression in breast cancer Cell Growth Differ 1997 8 773 778 9218871
Hajra KM Chen DY Fearon ER The SLUG zinc – finger protein represses E-cadherin in breast cancer Cancer Res 2002 62 1613 1638 11912130
|
16287501
|
PMC1308872
|
CC BY
|
2021-01-04 16:39:03
|
no
|
World J Surg Oncol. 2005 Nov 14; 3:73
|
utf-8
|
World J Surg Oncol
| 2,005 |
10.1186/1477-7819-3-73
|
oa_comm
|
==== Front
World J Surg OncolWorld Journal of Surgical Oncology1477-7819BioMed Central London 1477-7819-3-761630305610.1186/1477-7819-3-76Case ReportTeratoma occupying the left hemithorax Zisis Charalambos [email protected] Dimitra [email protected] Grigorios [email protected] Konstantinos [email protected] Konstantinos [email protected] Mihalis [email protected] Ion [email protected] Department of Cardiothoracic Surgery, Evangelismos General Hospital, Athens2 Department of Pathology, Evangelismos General Hospital, Athens3 Department of Critical Care and Respiratory disease, Evangelismos General Hospital, Ipsilantou 45-47, Athens2005 22 11 2005 3 76 76 13 7 2005 22 11 2005 Copyright © 2005 Zisis et al; licensee BioMed Central Ltd.2005Zisis et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Teratomas are manifested with a great variety of clinical and radiological features, while sometimes they simply represent incidental findings.
Case presentation
A rare case of benign teratoma of the dermoid cyst type, in an adult 40-year-old female patient, is reported. The patient had presented recurrent pulmonary infections for the previous 2 months, persistent cough, and progressively aggravating dyspnea. A chest X-ray showed total atelectasis of the left lung, and the thoracic CT-scan revealed a huge mass, containing multiple elements of heterogeneous density, probably originating from the mediastinum, occupying the whole left hemithorax. The mass compressed the vital structures of the mediastinum, great vessels and airways, and a chest MRI was performed to accurately detect the anatomical relations. The patient underwent left thoracotomy and the tumor was totally resected. The size of the tumor was extremely large although no invasion to the vessels or to the airway had occurred. Adherence to the adjacent left pulmonary artery and left main bronchus was present, but without erosion or fistulization. The postoperative course was uneventful, while the histological examination confirmed a teratoma.
Conclusion
A teratoma is a non-homogeneous pathological entity, clinically, radiologically or histologically. It is predominantly diagnosed between the second and fourth decade and the incidence is equal for both sexes. Symptoms are absent in one half of the patients. The case reported is noteworthy as the tumor appeared with total atelectasis of the left lung, and symptoms started 2 months prior to diagnosis. Total removal of the tumor is adequate treatment for this type of teratoma and the prognosis is excellent.
==== Body
Background
Mature teratomas are the most common histological type of germ cell tumors, followed by seminomas [1]. Germ cell tumors are predominantly found in gonads, while the anterior mediastinum is the most common extragonadal site [2]. The mediastinal germ cell tumors comprise 15% of anterior mediastinal tumors in adults and 25% in children [3]. According to the mediastinal germ cell tumor classification system proposed in 1986 by Mullen and Richardson, there are three categories: benign germ cell tumors, seminomas, and nonseminomatous germ cell tumors, also called malignant teratomas [4]. The benign germ cell tumors are also called epidermoid cysts, benign teratomas, or simply teratomas. These tumors can characteristically be cystic or solid or a combination of the two, contain multiple germ cell layers (sometimes all three are recognized, i.e., ectoderm, mesoderm, and endoderm), and are composed of tissue foreign to the organ or anatomic site in which they arise.
Case presentation
A 40-year-old female with negative smoking and medical history was admitted with productive cough, progressively aggravating dyspnea on exertion, and recurrent pulmonary infections for the previous 2 months. The chest X-ray showed total atelectasis of the left lung (Figure 1), and the thoracic CT-scan revealed a mass of the left hemithorax, which probably originated in the mediastinum and extended to the whole left pleural space (Figure 2). The mass showed heterogeneous density containing soft tissue elements, fat, cystic areas and foci of calcification, which is the classic imaging appearance of a benign teratoma on CT. Magnetic resonance imaging (MRI) was performed to specify the anatomic relationships and confirm the tumor morphologic features. The MRI yielded useful information about the vital structures of the mediastinum, whether invaded or externally compressed by the tumor. Specifically, the MRI confirmed a round, non-homogeneous, well circumscribed mass of a 12 cm diameter, exerting compression on the mediastinum great vessels and the left hilar structures (vessels and airway) (Figure 3). The bronchoscopy found stenosis of the trachea by external compression, and narrowing in the foramen of the left main bronchus resulting in difficulty in the insertion of the bronchoscope. The mediastinal tumor markers (α-fetoprotein and β-human chorionic gonadotropin) were both normal. As the findings of CT and MRI suggested a benign teratoma, a complete resection was contemplated. Since the mass was supposed to be respectable, surgical management was first in the priority list of therapeutic options. The cytological examination through transcutaneous needle aspiration or biopsy of the tumor were considered redundant and were omitted, because of the dispersion risk and the necessity for total removal so as to ameliorate the respiratory function and re-expand the left lung. Moreover, needle biopsy allows examination of only a small amount of tissue and may be inadequate for definitive diagnosis [5]. As already underlined in the literature, diagnosis and therapy rely on surgical excision, and even with large sized tumors whose complete resection is impossible, partial resection still relieves symptoms, frequently without relapse [6]. Pulmonary function tests were impaired: FVC 1.19 (36.7% of predicted) and FEV1 1.01 (41.5% of predicted), whereas a-FP, βCG were normal.
Figure 1 Preoperative chest X-ray showing total atelectasis of the left lung and deviation of the cardiac silhouette to the right.
Figure 2 Preoperative CT-scan of the chest revealing the mass compressing the mediastinal vessels and airway.
Figure 3 Preoperative MRI showing the teratoma, its anatomic relations to the mediastinum, and its expansion to the left hemithorax.
The patient underwent a total resection of the mediastinal mass via a left posterolateral thoracotomy. Entry into the pleural space was performed through the fifth intercostal space, and, because of the tumor large size, extension of the incision was necessary to obtain safe visualization of the cavity and proceed to tumor mobilization. Approaching via left thoracotomy makes access to the mediastinal structures difficult but permits control of the whole hemithorax up to the hilar structures. Many adhesions existed with the left pulmonary artery, the left main bronchus, the pericardium, the aorta, and the diaphragm, and a combination of blunt and sharp dissection for the division was applied uneventfully. Because of difficulty in the mobilization of such a huge mass, a purse string suture permitted aspiration of sebaceous content via a small incision in the wall. As the size diminished, manipulation was facilitated. The tumor, excised en block, was white-gray colored, well circumscribed, and thick capsuled. A tube thoracostomy was introduced, the collapsed left lung was easily re-expanded, and the patient was extubated. The patient recovered well from the operation and was discharged on the 2nd postoperative day. Preoperative atelectasis of the left lung was totally resolved (Figure 4), and the pathological examination revealed a benign mature teratoma with dermoid cyst characteristics, containing sebaceous and gelatinous material. Two years later, the patient is doing well out of recurrence.
Figure 4 Postoperative X-ray of the chest after total resection of the teratoma, showing re-expansion of the left lung.
Discussion
Germ cell tumors typically occur in young adults in their second to fourth decade with equal sex distribution. Female predominance has been reported by some authors with a 1.27–2.05: 1 female: male ratio [7-9]. Multicompartment extension is observed in 10–15% of the cases [2]. According to another review, 3–8% are located in the posterior portion of the visceral compartment or the paravertebral regions [10,11], where neurogenic tumors and neoplastic lymphadenopathy commonly exist. In the case reported, the teratoma occupied at least two of the three mediastinal compartments (the anterior and the middle) exerting compression on relevant vital anatomical structures.
Nearly one half of patients (36–62% in various series) have no signs or symptoms when the mass is initially diagnosed. Symptoms commonly present are chest, back or shoulder pain, dyspnea, cough, fever, pleural effusion, and bulging of the chest wall [7]. In the case under discussion, the patient presented symptoms from her tumor for 2 months, either due to tumor inflammation or the total atelectasis of the lung, following recurrent pulmonary infections. Such a sizeable tumor was thus asymptomatic for a long period. Hemoptysis or expectoration of hair or sebum can rarely occur when communication between the tumor and the tracheobronchial tree develops. Symptoms can also derive from the pressure exerted on the surrounding tissues (for example superior vena cava syndrome).
In spite of the pressure on the mediastinal vessels or the airway, invasion of these anatomical elements does not usually occur. However, erosion of the tumor to the bronchus, as well as to the pleural space, to the skin, and to the aorta has been reported.
Diagnostic assessment is performed with classical X-rays, followed-up by CT. Typically, a well-circumscribed anterior mediastinal mass extending to one side of the midline and protruding into one lung field is revealed. CT accurately estimates the density of all included tissues, such as soft tissue (in virtually all cases), fluid (88%), fat (76%), calcification (53%), and teeth and such imaging findings are considered specific [2]. MRI is valuable in detecting the anatomic relations to the mediastinal and the hilar structures, such as vessels and airways [12].
Macroscopically the tumors are spherical, lobulated, with a well-defined capsule, and contain a variety of material, lipid-rich fluid, cheese-like substances, teeth, hair, and cartilage. Histologically, benign teratomas comprise at least 2 of the 3 primordial layers: ectoderm (skin and hair), mesoderm (bone, fat, and muscle), and endoderm (respiratory epithelium and gastrointestinal tract).
Surgical resection is the treatment of choice and radical extirpation secures a long survival out of recurrence. Median sternotomy is usually preferred for tumor removal, but access via either posterolateral or anteroposterior thoracotomy depends on the size, location, and expansion of the tumor. Difficulty in surgical maneuvers may be a result of the vital structures involved. In a large series of 95 patients with benign mature teratoma, in addition to tumor resection, 3 patients required lobectomy, 5 patients additional partial resection of the lung, and 7 patients pericardectomy. VATS techniques have been introduced in teratoma resection with promising results [7,13]. The appropriate operative procedure choice, depending on the patient's age, tumor size, location, expansion, psychological profile, coexistent morbidity, cardiorespiratory reserve, and preference, as well as the surgeon's experience guarantee a favorable outcome and long-term prognosis.
In the case reported, left posterolateral access was chosen, as the tumor was almost entirely located into the left hemithorax, reaching and adherent to the left hemidiaphragm. Such a location precludes median sternotomy, otherwise the preferred approach, as surgical manipulations are impossible on the lower lobe of the left lung and on the left hemidiaphragm. MRI was performed after the CT-scan in order to more precisely reveal the relations with the mediastinum vital anatomical structures (major vessels, aorta, pulmonary artery) because it was critical to clarify whether there was an invasion or simply compression. The examination pointed to the latter and loose tumor adhesions with the great vessels were confirmed on surgery and gently submitted to blunt dissection. The left lateral decubitus position of the patient did not influence ventilation during operation, as ventilation conditions with a double lumen tracheal tube were similar to the preoperative values affected by the left lung atelectacis. Furthermore, left lung operation secured the right lung against potential contamination from infected content of the chronically collapsed left lung.
Conclusion
• Such a huge teratoma extending to the whole hemithorax and resulting in total atelectasis of the left lung has been only sparsely reported previously [14,15].
• Even a huge mass causes simple compression and deviation without invasion of the vascular and airway structures.
• Complete resection is adequate treatment for the patient with favorable long-term prognosis.
Conflict of interest
The author(s) declare that they have no competing interests.
Authors' contributions
CZ was the main surgeon of the patient, who reviewed the literature and had the central responsibilityin the management ofthe patient and the writing of this report,
DR was the pathologist, who examined and diagnosed histologically the resected specimen,
GS was the interventional bronchoscopist and pulmonologist, who performed bronchoscopy, functional respiratory tests and assessed the patient as candidate of major surgical procedure,
KV and KS were assistants in this operation and helped in preparation of the manuscript,
MA offered his valuable experience in the management of the patient and the writing of this report,
IB coordinated writing of the manuscript.
All authors read and approved the final manuscript.
Acknowledgements
Written consent of the patient was obtained for publication of his case report.
==== Refs
Nichols CR Mediastinal germ cell tumors. Clinical features and biologic correlates Chest 1991 99 472 479 1846573
Moeller KH Rosado-de-Christenson ML Templeton PA Mediastinal mature teratoma: imaging features AJR Am J Roentgenol 1997 169 985 990 9308448
Rosado-de-Christenson ML Templeton PA Moran CA From the archives of the AFIP. Mediastinal germ cell tumors: radiologic and pathologic correlation Radiographics 1992 12 1013 1030 1326777
Mullen B Richardson JD Primary anterior mediastinal tumors in children and adults Ann Thorac Surg 1986 42 338 345 3530162
Allen MS Trastek VF Pairolero PC Thomas W Shields, Joseph LoCicero III, Ronald B Ponn Benign germ cell tumors of the mediastinum General Thoracic Surgery 2000 2 5 Lippincott Williams & Wilkins 2281
Davis RD JrOldham HN JrSabiston DC Jr Sabiston, Spencer The Mediastinum Surgery of the chest 1996 I 6 WB Saunders 596
Takeda S Miyoshi S Ohta M Minami M Masaoka A Matsuda H Primary germ cell tumors in the mediastinum. A 50-year experience at a single Japanese institution Cancer 2003 97 367 376 12518361 10.1002/cncr.11068
Lewis BD Hurt RD Payne WS Farrow GM Knapp RH Muhm JR Benign teratomas of the mediastinum J Thorac Cardiovasc Surg 1983 86 727 731 6632945
Dulmet EM Macchiarini P Suc B Verley JM Germ cell tumors of the mediastinum. A 30-year experience Cancer 1993 72 1894 1901 7689921
Shirodkar NP Chopra PS Marker M Murphy KD Dhamoon A Kwon OJ Conjoined gastric and mediastinal benign cystic teratomas. Case report of a rare occurrence and review of literature Clin Imaging 1997 21 340 345 9316754 10.1016/S0899-7071(96)00078-2
Sinclair DS Bolen MA King MA Mature teratoma within the posterior mediastinum J Thorac Imag 2003 18 53 55 10.1097/00005382-200301000-00010
Drevelegas A Palladas P Scordalaki A Mediastinal germ cell tumors: a radiologic-pathologic review Eur Radiol 2001 11 1925 1932 11702124 10.1007/s003300000725
Cheng YJ Huang MF Tsai KB Video-assisted thoracoscopic management of an anterior mediastinal teratoma: report of a case Surg Today 2000 30 1019 1021 11110399 10.1007/s005950070025
Sohn L Gribbin C Rizzo N Nosher JL Radiology/pathology conference at Robert Wood Johnson Medical School. Benign mediastinal teratoma N J Med 1995 92 241 244 7746517
Tominaga K Kadokura M Saida K Nakao K Kushiro H Ryu K Yamamoto N A surgical case of giant mediastinal teratoma Kyobu Geka 1994 47 944 947 7967269
|
16303056
|
PMC1308873
|
CC BY
|
2021-01-04 16:39:03
|
no
|
World J Surg Oncol. 2005 Nov 22; 3:76
|
utf-8
|
World J Surg Oncol
| 2,005 |
10.1186/1477-7819-3-76
|
oa_comm
|
==== Front
BMC Evol BiolBMC Evolutionary Biology1471-2148BioMed Central London 1471-2148-5-621627766010.1186/1471-2148-5-62Research ArticleUnusual linkage patterns of ligands and their cognate receptors indicate a novel reason for non-random gene order in the human genome Hurst Laurence D [email protected] Martin J [email protected] Department of Biology and Biochemistry, University of Bath, Bath, BA2 7AY, UK2005 8 11 2005 5 62 62 14 6 2005 8 11 2005 Copyright © 2005 Hurst and Lercher; licensee BioMed Central Ltd.2005Hurst and Lercher; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Prior to the sequencing of the human genome it was typically assumed that, tandem duplication aside, gene order is for the most part random. Numerous observers, however, highlighted instances in which a ligand was linked to one of its cognate receptors, with some authors suggesting that this may be a general and/or functionally important pattern, possibly associated with recombination modification between epistatically interacting loci. Here we ask whether ligands are more closely linked to their receptors than expected by chance.
Results
We find no evidence that ligands are linked to their receptors more closely than expected by chance. However, in the human genome there are approximately twice as many co-occurrences of ligand and receptor on the same human chromosome as expected by chance. Although a weak effect, the latter might be consistent with a past history of block duplication. Successful duplication of some ligands, we hypothesise, is more likely if the cognate receptor is duplicated at the same time, so ensuring appropriate titres of the two products.
Conclusion
While there is an excess of ligands and their receptors on the same human chromosome, this cannot be accounted for by classical models of non-random gene order, as the linkage of ligands/receptors is no closer than expected by chance. Alternative hypotheses for non-random gene order are hence worth considering.
==== Body
Background
One of the most striking discoveries in the post-genomic age has been the amount of non-random gene positioning in eukaryotic genomes [1]. In the human genome, for instance, highly/broadly expressed genes cluster [2-5]. Likewise in yeast co-expressed genes tend to reside together [6] and such pairs tend also to be retained together over evolutionary time more than expected, given the intergene distance between them [7]. Blocks of broadly expressed mammalian genes also seem to be preserved over evolutionary time more than expected [8]. In Caenorhabditis [9-11], Drosophila [12-14] and Arabidopsis [15], to name but three, there exists further evidence for expression clusters of some variety. These results all suggest that eukaryotic genomes are organised in a manner that permits co-expression or co-ordinate expression. Evidence also suggests linkage of functionally related genes, although on this issue the evidence is more equivocal, not least because of an ambiguity as to what "functionally related" can mean. On the one hand, in numerous eukaryotic genomes, genes from the same metabolic pathway cluster more than expected by chance [16] (for detailed case history see [17]). Likewise, linked co-expressed genes in yeast often fall within the same MIPs (Munich Information Centre For Protein Sequences) category [6] or the same Gene Ontology (GO) classification [18].
These results are so striking because they so profoundly overturn the long held assumption that genes are randomly located around eukaryotic genomes. This is not to say that possible exceptions were not considered prior to the sequencing of the complete genome. They were, however, typically dismissed as being unrepresentative or uninteresting either because they were clearly the product of tandem duplication (hox cluster, globin cluster) or were associated with weird genetics (imprinted clusters) or genes that are otherwise exceptional (e.g. clustering of rRNAs). Not all such suggestive examples could so easily be dismissed however. Here we concentrate on one class, linkage of ligands to their cognate receptors. This issue is worth systematic analysis, not least because in yeast it has recently been shown that genes whose proteins interact to form stable complexes are linked more often than expected by chance [19].
That ligands and their receptors may be linked was observed independently by several workers. Cooper [20], noting that the linkage of ligands to receptors may be common, highlights the examples of transferrin and transferrin receptor on chromosome 3q, as well as apolipoprotein E and the low density lipoprotein receptor both on chromosome 19. He also rightly cautions, however, than one can find numerous cases where ligands and receptors are not linked. Similarly, Lennard et al. [21] note the linkage of the three ligands in the interleukin 1 cluster (IL1 alpha, beta and receptor antagonist) to the two receptors [21]. The linkage of ligands to receptors has even proven to have some predictive power. Wang et al. [22] noticed that hepatocyte growth factor (HGF) and its MET receptor were both on 7q. Noting too the presence in 3p21 of both macrophage stimulating factor (MST1, a member of the same gene family as HGF), and RON (a member of the MET receptor family), they hypothesised that RON might be MST1's receptor [22]. This, in turn, they demonstrated to be the case (RON's alias is now MST1R) [22]. Popovici et al. note several of the above examples and also point to a total of 14 incidences of linked genes involved in the same pathway, not necessarily as ligand-receptor couplings [23].
This evidence prompts two questions. First, is it true that there is something odd about the linkage patterns of ligands and their cognate receptors? Second, if it is true, why might this be so? Prior authors have also suggested that linkage of ligands to receptors might be functionally important. Haig [24], observing two of the above cases (interleukin 1 and transferrin), notes that close proximity could enable linkage disequilibrium between alleles at the ligands and receptors. This linkage disequilibrium would potentially enable the spread of rare allele combinations for which there exist particular epistatic interactions. These may, Haig suggests, act as selfish maternal effect lethals, an example of which has been described in mice [25,26]. This theory may be seen as a special case of a more general theory for linkage based on preservation of linkage disequilibrium under epistasis [27-30]. One might also conjecture that ligands and receptors might at times need to be co-expressed [see e.g. [31]], so very close linkage might be beneficial for this reason as well.
If selection does act on the location of ligands and receptors (either to permit co-expression or to maintain linkage disequilibrium), then we should predict, from the above models, that when two such genes are on the same chromosome they should also be, on average, physically closer than would be expected by chance. To this end we ask two questions. First, is the mean distance between ligands and their linked receptors shorter than expected by chance? As this mode of analysis could miss an excess of cases with very tight linkage, we additionally ask whether the number of incidences of linkage within a given window size (1 Mb, 2 Mb etc.) is higher than expected by chance.
Results and discussion
No evidence for close proximity of ligands and receptors
If ligands and their cognate receptors were under selection to be in close physical proximity, we should find that the mean distance between them should be smaller than expected by chance. To test this, we examined ligand-receptor pairs from the DLRP database [32,33]. All analyses were also performed for an augmented dataset, which additionally contains two 'cherry-picked' cases highlighted by Cooper [20] (see Methods). Contrary to expectations, the mean distance between ligand and receptor in the non-augmented data set (64.579 Mb) is higher than that found in randomized genomes (56.344 Mb; P = 0.733). The same pattern is found in the augmented data set (real = 63.466 Mb, randomized = 56.165 Mb, P = 0.709). Note that a ligand can have many receptors and that this is factored into the analysis through the randomization protocol.
It may, however, be the case that there exist a number of ligand-receptor pairs that are much closer to each other than expected. To examine this we compared the number of ligand-receptor pairs within some critical distance of each other and compared this with the number expected by chance. In these simulations we permuted genes only within the chromosomes within which they are found, so as to ensure that the number of ligand-receptor pairs were the same in the randomized sets as in the real data set. As can be seen (Table 2) in neither data set do we find evidence for anything other than a pattern of random linkage [see Additional file 4].
Table 2 The number of occurrences of a ligand receptor pair within a given distance of each other (in Mb) for the real genome compared with 10000 randomized genomes. In this instance the first three results columns refer to the dataset excluding the two examples identified by Cooper.
Distance (Mb) Obs Exp P Obs Exp P
0.2 0 0 1 0 0 1
0.5 1 0.09 0.095 1 0.09 0.087
1 1 0.17 0.163 1 0.16 0.152
2 1 0.94 0.672 1 0.94 0.67
5 1 1.62 0.864 1 1.59 0.855
10 2 3.18 0.927 2 3.16 0.917
20 7 6.20 0.413 7 6.18 0.409
50 12 10.94 0.363 13 12.0 0.385
It has been previously established that broadly expressed genes cluster [3,4]. Might our simulations have produced misleading results by permitting genes of the ligands and receptors to reside in any chromosomal location? To examine this possibility we considered randomizations in which genes are swapped exclusively with ones of the same breadth of expression. None of the above results are qualitatively affected (Table 3). Using a different bin size to classify breadth of expression appears to have no effect on the results (Table 3). Likewise, permitting ligands and receptors to be located in the genome at the locations of other ligands and receptors does not affect any conclusions [see Additional file 4].
Table 3 Results of randomizations controlling for breadth of expression. The "Aug" (Augmented) data set is that containing the two ligand -receptor sets nominated by Cooper [20]. Bin size indicates the span of breadths of expression considered to be the same in the randomizations. Bin size one implies that only genes of the same breadth were switched with each other. Distances are measured in Mb. P-values are estimated by comparison of observed data with expectations obtained from randomised genomes.
Data set Bin size N # on same chr (observed) # on same chr (expected) P Mean dist. (observed) Mean dist. (expected) P
Aug 1 267 25 14.18 0.011 63.47 55.22 0.741
Aug 5 267 25 14.14 0.01 63.47 55.77 0.705
Non-Aug 1 265 23 14.09 0.027 64.58 56.25 0.731
Non-Aug 5 265 23 13.8 0.013 64.58 55.47 0.752
The above results indicate that there is no evidence for selection for clustering of ligand and receptor. For this reason we reject a model positing epistasis between alleles of ligand and receptor as a general force acting on genomic location of these genes. Moreover the lack of tight clustering suggests that we are not witnessing clustering to enable co-regulation (by ensuring that genes are co-localised in the same chromatin block). The model suggested by Haig [24] is not, however, necessarily falsified by the above results, as he postulates selection on disequilibrium only if the genes might be involved in maternal-foetal interactions. Such a model is hard to falsify in the absence of segregation/viability data from appropriate haplotypes. However, we can note that if we further restrict our data sets to those in which either the ligand or one of the receptors is placentally expressed, the qualitative patterns described above are unaltered [see Additional file 4]. We find, therefore, no evidence for close linkage of ligand and receptor when involvement in maternal-foetal interactions might be a possibility.
An excess of ligand-receptor pairs on the same human chromosome
Above we asked whether ligands and their receptors are more closely linked than expected by chance. We can also ask if ligands and their receptors are more commonly linked (i.e. on the same chromosome) than expected by chance? In an unbiased unaugmented human data set (i.e. without the addition of the two sets highlighted by Cooper [20], see Methods) we observe 23 such pairings but expect on average 13.71 (P = 0.015). When we include the two extra sets the P value, as expected, is reduced: we observe 25 pairs but expect on average 13.8 (P = 0.005). These results support the view that in the human genome linkage of a ligand to at least one of its cognate receptors is more common than would be expected by chance. However, the majority (approx 78%) of ligands are not linked to any of their receptors, so this excess should not be considered a strong rule (although, as already noted, in special cases it has had predictive power).
No evidence for an excess of ligands-receptor pairs on the same chromosome in mouse
To ask whether the patterns observed in the human genome are also found in the mouse genome we constructed three mouse data sets and applied the three randomization protocols to each. The first two data sets are the ortholog equivalents of our two human data sets purged of duplicates by either a) Blasting or b) Blasting and removal by physical proximity (in the human genome) of ligands or receptors. That is, if two ligands were in close proximity in the human genome, even if not identified as sequence related, we would remove one before considering the location of the mouse orthologs. However, as it is possible that some ligand clusters might be unique to mouse, we additionally purged the more stringent of the above two of any groupings of ligands or receptors seen in the mouse genome.
As it happens, no matter which data set one employs or which randomization method is performed, there is not even a remote hint that ligands and cognate receptors occur more commonly on the same mouse chromosome than expected by chance [see Additional file 5]. For example, in the equivalent of the human data set purged of duplicates by Blast alone, we observe 19 ligand-receptor pairs on the same chromosome and expect 19.38 (P = 0.56) in a randomization in which the ligands and receptors can assume any genomic position currently associated with a gene. In this analysis the mean distance between ligand and receptor is 57.3 Mb but is 43.4 in the randomizations (P = 0.937). At no specified distance do we find more pairs than expected by chance [see Additional file 5]. Controlling for breadth of expression makes no difference to this conclusion [see Additional file 6]. The list of linked genes from the data set equivalent to the human set presented in Table 1 is presented in Table 4. In this set 16 ligand-receptor pairs co-occur on the same chromosome, with 14.5 expected. Only four ligand-receptor pairs are in common in the two comparable data sets.
Table 1 Incidences of occurrence, on the same human chromosome, of a ligand with one of its cognate receptors after removal of tandem duplicates by Blast and by the physical proximity method. The distance is defined as the span between the mid-positions of the ligand and the mid-position of the receptor.
Class Gene Unigene # Cytogenetic Distance (Mb)
Ligand DLL1 368657 6q27
Receptor NOTCH4 436100 6p21.3 138.23
Ligand DLL3 127792 19q13
Receptor NOTCH3 8546 19p13.2-p13.1 29.53
Ligand EFNA4 449913 1q21-q22
Receptor EPHA8 283613 1p36.12 129.2
Ligand TNFSF18 248197 1q23
Receptor TNFRSF18 212680 1p36.3 168.57
Ligand IL1RN 81134 2q14.2
Receptor IL1R2 25333 2q12-q22 11.51
Ligand HGF 396530 7q21.1
Receptor MET 419124 7q31 34.96
Ligand MST1 349110 3p21
Receptor MST1R 2942 3p21.3 0.21
Ligand FGF1 278954 5q31
Receptor FGFR4 165950 5q35.1-qter 34.45
Ligand FGF2 422889 4q26-q27
Receptor FGFR3 1420 4p16.3 122.37
Ligand FGF8 57710 10q24
Receptor FGFR2 404081 10q26 19.77
Ligand FGF10 248049 5p13-p12
Receptor FGFR4 165950 5q35.1-qter 132.07
Ligand FGF17 248192 8p21
Receptor FGFR1 748 8p11.2-p11.1 16.46
Ligand FGF18 87191 5q34
Receptor FGFR4 165950 5q35.1-qter 5.65
Ligand VEGFC 79141 4q34.1-q34.3
Receptor KDR 12337 4q11-q12 122.23
Ligand PTN 44 7q33-q34
Receptor PTPRZ1 78867 7q31.3 15.23
Ligand TGFB2 169300 1q41
Receptor TGFBR3 342874 1p33-p32 122.99
Ligand BMP3 121507 4p14-q21
Receptor BMPR1B 480321 4q22-q24 13.97
Ligand BMP10 158317 2p13.3
Receptor ACVR1 150402 2q23-q24 89.45
Receptor ACVR2 283349 2q22.2-q23.3 79.47
Ligand INHBB 1735 2cen-q13
Receptor ACVR1 150402 2q23-q24 37.64
Receptor ACVR2 283349 2q22.2-q23.3 27.66
Ligand INHA 407506 2q33-q36
Receptor ACVR1 150402 2q23-q24 61.8
Receptor ACVR2 283349 2q22.2-q23.3 71.79
Ligand TF 433923 3q22.1
Receptor TFRC 185726 3q29 62.32
Ligand APOE 110675 19q13.2
Receptor LDLR 213289 19p13.3 39.02
Table 4 Incidences of occurrence, on the same mouse chromosome, of a ligand with one of its cognate receptors after removal of tandem duplicates by Blast and position methods. The distance is defined as the span between the mid-positions of the ligand and the mid-position of the receptor. The receptors indicated with a Y in the conserved linkage column are those that are also on the same chromosome as the same ligand in the comparable human genome set (i.e. those in Table 1).
Class Gene Unigene # Chromosome Distance (Mb) Conserved linkage
Ligand Jag1 Mm.22398 2
Receptor Notch1 Mm.290610 2 110.41
Ligand Dll4 Mm.143719 2
Receptor Notch1 Mm.290610 2 92.67
Ligand Bmp2 Mm.103205 2
Receptor Acvr1 Mm.689 2 74.85
Receptor Acvr2 Mm.314338 2 84.49
Ligand Bmp7 Mm.595 2
Receptor Acvr1 Mm.689 2 114.51
Receptor Acvr2 Mm.314338 2 124.15
Ligand Tnfsf8 Mm.4664 4
Receptor Tnfrsf8 Mm.12810 4 81.45
Ligand Pdgfa Mm.2675 5
Receptor Pdgfra Mm.221403 5 62.52
Ligand Ptn Mm.279690 6
Receptor Ptprz1 Mm.41639 6 13.86 Y
Ligand Fgf15 Mm.3904 7
Receptor Fgfr2 Mm.16340 7 14.87
Ligand Mst1 Mm.8369 9
Receptor Mst1r Mm.3901 9 0.17 Y
Ligand Ifng Mm.240327 10
Receptor Ifngr1 Mm.549 10 98.83
Ligand Fgf10 Mm.317323 13
Receptor Fgfr4 Mm.276715 13 61.47 Y
Ligand Bmp4 Mm.6813 14
Receptor Bmpr1a Mm.237825 14 8.55
Ligand Dll1 Mm.4875 17
Receptor Notch3 Mm.4945 17 16.56
Receptor Notch4 Mm.173813 17 18.88 Y
Explaining the data: interesting biology or statistical artefact?
We find no evidence in mouse or man that ligands and their receptors are more closely linked on average than expected by chance. However, we do find that there are more ligand-receptor pairs on the same human chromosome than expected, a feature not found in mouse. From the above results we are faced with two possible explanations for the human data. First, that it is just a statistical blip, possibly owing to some subtle bias in the original data set (note for example, that MST1R was analysed as a potential receptor because of its linkage to MST [22]). Second, that the excess of ligand-receptor pairs on the same chromosome is the product of some deterministic force, that for some reason does not apply, or is not strong enough, in rodents. This might either be direct selection favouring the persistence of co-occurrence or a deterministic bias in the creation of co-occurrence.
That the pattern is found in humans rather than mice argues against a direct selective benefit for co-retention on the same chromosome. This is owing to the fact that the effective population size of the human population is most probably much smaller than that of mice. As such, according to the nearly-neutral theory, the efficacy of selection should be higher in mice (see also [34,35]). Hence, if the pattern was owing to selection directly favouring co-occurrence, it is more likely to be observable in mouse rather than human, all else being equal. Moreover, it is also hard to see what direct selective benefit might accrue from co-occurrence in weak linkage. In particular, co-regulation of ligands and receptors (which is not clearly expected in the first place) would likely require much tighter linkage than observed here. We also find no evidence for a stronger similarity in the breadth of expression of the linked ligands and receptors than those on different chromosomes. We considered the difference in breadth of expression between ligand and receptor normalised by the mean of the two. Linked genes are of no more similar breadth of expression (mean difference for linked genes 0.75 +/- 0.12, for unlinked 0.686 +/- 0.036, t-test, P = 0.58).
Segmental duplication and the balance hypothesis: an alternative hypothesis for non-random gene order
A notable feature of our data set is that there are numerous cases in which a ligand-receptor pair in linkage is matched by at least one other paralogous pair also in linkage. If we define genes belonging to the same Hovergen [36] family as paralogs, then we can identify the following linked paralogous pairs from Table 1: HGF/MST1 (ligands) with MET/RON (receptors); DLL1/DLL3 with NOTCH3/NOTCH4; FGF1/FGF2 with FGFR3/FGFR4; FGF8/FGF17 with FGFR1/FGFR2 (see also [23]). Note too that FGF18 is linked to FGFR4 and to FGF1 and is sequence related to FGF8 and FGF17.
It has been argued that co-paralogy of gene pairs involved in the same pathway (of which ligand-receptor pairs are but one example) appear to be unusually common [23]. This finding is also in accord with much recent evidence suggesting that the human genome (and the vertebrate genome more generally) may be a mosaic of old large block duplications (i.e., duplications of large chunks of DNA sequence) and/or the result of whole genome duplications [37], with several of the above paralogous groups being claimed to be the result of such duplication events: FGFR 1, 2, 3 and 4 are in paralog clusters on human chromosomes 8p, 10q, 4p, and 5q respectively [38]; NOTCH 3 and 4 also appear to be in paralog clusters on chromosomes 6 and 19, with NOTCH 1 and NOTCH 2 being two further duplications of the same block [39]; HGF/MST1 and MET/RON were also previously described as belonging to co-paralogous groups [23].
Following an earlier hypothesis [19], we would like to suggest an hypothesis to explain our results that is based on the occurrence of block duplications [40]. If some ligands and receptors require an appropriate balance in their titres, then one could expect that a mutation resulting in a block duplication containing one of the pair (e.g., the ligand) might be more likely to spread through the population if it also duplicates the other (e.g., the receptor). Such co-duplication is most likely if ligand and receptor happen to be linked, while unlinked ligand-receptor pairs are less likely to be successfully duplicated. Our hypothesis may be considered as being a form of the balance hypothesis [41,42], which supposes that proteins involved in mutual interactions need to have their titres appropriately balanced. Direct evidence for this proposition has been described in yeast, in which it is also reported that the need for balance might explain the lack of duplicability of the genes involved in complex formation [43].
This hypothesis appears tenable in the current context for several reasons. It is, for example, suggestive that the paralogous pair sets tend to be more closely linked, although the statistic is on the edge of significance (Median Test, ChiSquared = 3.59, P = 0.058, df = 1). This would be expected if constraints exist on the upper size limit of the block duplications. It may also be notable that for the many of the above genes there is evidence for dosage sensitivity, as required by the balance hypothesis (Table 5). Moreover, as re-arrangements tends to be especially common in mice [44-46], any linked pairs possibly generated by block duplication are more likely to be split up, making the genome more like random.
Table 5 Evidence for or against dosage sensitivity of ligand and receptors that were possibly the source for or the consequence of block duplication
Gene Evidence for dose sensitivity References
HGF Over-expression in retinal pigment epithelium induces retinal detachment [58]
MET* Autosomal dominant Hereditary papillary renal carcinoma is associated with mutations in MET [59]
MST1 None: mouse knockout is without strong phenotype [60]
RON Hemizygous mice (Ron +/-) are highly susceptible to endotoxic shock and are compromised in their ability to downregulate nitric oxide production [61]
DLL1 No report of heterozygous null phenotype nor of overexpression phentype
NOTCH3* Autosomal dominant disorder CADASIL owing to mutation in NOTCH3 [62]
DLL3 None: the gene is associated with disease (SCDO1/2) in mutant homozygotes but no report of heterozygote phenotype.
NOTCH4 Upregulation of NOTCH4 is associated with mammary tumours See OMIM 164951
FGF1 None: no phenotype in Fgf1 homozygous knockouts [63]
FGFR3* Autosomal dominant disorder ACH associated with mutation in FGFR3 See OMIM 134934
FGF2 Over-expression promotes bone growth [64]
FGFR4 None: knockout homozygotes have no obvious phenotype [65]
FGF8 Over-expression is associated with carcinogenesis [66]
FGFR1* Autosomal dominant Pfeiffer syndrome is owing to mutations in FGFR1 [67]
FGF17 None: heterozygote knockout has no phenotype [68]
FGFR2* Numerous autosomal dominant disorders associated with FGFR2 See OMIM 176943
FGF18 Dose sensitive liver and small intestine development [69]
*Note: These five genes are known to be associated with human autosomal dominant disorders. Searching OMIM [70] for mapped genes with "autosomal dominant" somewhere in the title or the text and not "recessive" in the title, reveals an estimate of 1166 mapped genes which may be associated with autosomal dominant disorders. Assuming 22470 autosomal genes (as in the NC files) we should then have expected from a random sample of 17 autosomal genes less than one dominant, significantly less than observed (Chi-squared, P < 0.0001). However, in some cases the dominance is owing to negative mutations rather than dose per se.
The above hypothesis also predicts that ligands and their linked receptors might be duplicated at the same time. While this can be approximately established by phylogenetic methods, these do not constitute a perfect test, as they fail to establish whether the pairs were duplicated in a block together and furthermore, genes known to be co-duplicated are very commonly not identified as such by phylogenetic methods [47]. Nonetheless, we have surveyed the available data and prior analyses and fail to find any data that contradicts the hypothesis that when a given receptor duplicated the relevant ligand did as well [see Additional file 7]. The ligands FGF1 and FGF18 are, however, probably the result of an ancient duplication that occurred independent of the receptor [48].
In some of the incidences reported here, a case for co-duplication has already been made. The linkage of the FGFs to their receptors has previously been argued, from phylogenetic data, to be owing to block or whole genome duplications [48]. This view is supported by our inspection of the phylogenetic tree of the FGFR family as presented in Hovergen [36], which suggests that at the base of the vertebrates there was one receptor which duplicated to produce the ancestors of FGFR1/2 and FGFR3/4. Duplication of both ancestral sequences then occurred very shortly after (prior to the divergence of the fish), leaving FGFR1 and FGFR2 as nearest paralogs, and FRGR3 and FGFR4 as nearest paralogs. If there was co-duplication of the receptors, we should expect to see FGF1 and FGF2 as nearest paralogs and FRF8 and FGF17 as nearest paralogs, with, in both incidences, duplication occurring near the base of the vertebrates. The nearest paralog relationships are indeed upheld [48]. Furthermore, in both instances the duplication occurred prior to the divergence of fish, as predicted.
Conclusion
In sum, we have described a novel pattern of co-localisation on the same chromosome of genes whose products interact, which cannot obviously be accounted for either by known models for co-ordinate regulation, nor by selection for linkage disequilibrium. The pattern may in part reflect a past history of block duplication. A version of the balance hypothesis is worth considering as underpinning to explain the results.
Tests of this hypothesis should be possible in the future. We should in principle be able, with fuller knowledge of gene order in many mammals, to reconstruct the past history of duplication and gene order re-arrangements that occurred through mammalian history. The model predicts an excess of block duplications in which both ligand and cognate receptor are found, as well as excess in which neither are found, but a dearth of those with one, but not the other. The model also predicts a general weakening of this initial signal with increasing numbers of inter-chromosomal re-arrangements, as the hypothesis proposes only an initial filter of block duplications, not ongoing direct selection to maintain linkage.
Methods
Data set assembly and curation
The table of ligand-receptor partners were extracted from the DLRP database [32,33]. This specifies for any given single ligand the corresponding receptor or set of receptors. The genes here are referred to by gene name and Unigene id number, by reference to an old release of Unigene. These entries we updated to the current release for Homo sapiens, UniGene Build #175. For each Unigene number and gene name, the relevant Unigene page was identified [49]. If the entry remained in the new build all details were left unchanged, except in three cases where there exists a gene by the same name as that in the original dataset, at the same genomic location as the given Unigene entry, but in a separate Unigene class. In these cases the Unigene entry with matching name was employed. If the old entry had been retired then a) if only one new entry is available this was used, b) if multiple entries were found (i.e. the cluster has split), then the one with the gene name identical to that of the old entry was used, c) if no entry had the same name but all entries were at the same genomic location the entry with the most abundant sequence data was used, d) if no unambiguous match could be found the entry was eliminated. If this was the ligand then the whole entry was deleted. In a few cases separate ligand receptor blocks are collapsed to the same Unigene entry in build #175 (e.g. FGFR and FGFRB). In this instance one of the two sets was eliminated. The original file also contains a number of entries in which the ligand alone is given, with no receptor. In these cases the entry was deleted.
From this new set of Unigene identities Entrez gene [50] was searched with the current Unigene id being posted. From here we recovered a) the Entrez/LocusLink gene name b) the Entrez/LocusLink id. If the LocusLink/Entrez gene name was different from the Unigene name then the pairs were examined at LocusLink to determine that the names were synonymous. In all cases this proved to be so. From here we obtained the physical location in the NC Genbank files for each chromosome. The cDNA source annotation of each Unigene entry was employed to determine whether the gene was placentally expressed.
The data set does not include the two cases highlighted by Cooper [20]: transferrin and its receptor and apolipoprotein E and its receptor. We therefore consider a second data set in which we add these two. This is, however, problematic as we are adding only "cherry picked" data. It is thus much more likely that we should find close linkage in the expanded set compared with the original set, owing to the non-random nature of the addition to the data set. Nonetheless, should we find an absence of an effect in this expanded set, this would make for stronger evidence against the hypothesis of an over-abundance of close linkage of ligands and their receptors.
The set so defined has numerous clusters of sequence-related ligands and receptors. Such clusters are likely to have arisen from tandem gene duplications, and thus individual genes cannot be treated as independently positioned. To eliminate the effects of tandem duplication we perform an all versus all blast (with E < 0.01) of the coding sequences defined from the RefSeqs for each gene. For each pair of putative duplicates on the same chromosome one of the two was randomly selected to be removed. In a tandem cluster with more than two duplicates only one gene was considered. Using E < 0.2 resolves to the same data set.
With this approach, however, a few well described duplicate clusters are not identified. For example, there remain 6 ligand-receptor pairs associated with the 2q14 cluster of three ligands (IL1 alpha, beta and the receptor antagonist) and their two receptors (IL1R1 and Il1R2) in 2q12. However, while Blast fails to reveal either of these clusters as duplicate clusters, this contradicts the conventional wisdom, based on close analysis of gene structure, function and conserved functional parts, that they are both duplicate arrays [51,52]. The problem in this instance is most probably that interleukins and their receptors tend often to be fast evolving, hence liable to avoid detection as duplicates unless they are relatively modern duplicates. Indeed this inability of Blast to identify orthology/paralogy of fast evolving genes has recently been well demonstrated [53].
To eliminate such problems we additionally remove one of a pair of ligands within 1 Mb of each other. Likewise we remove one of a pair of receptors should the receptors occur within 1 Mb of each other. In nearly all cases the ligands in the cluster also bind the same receptors and vice versa. One exception is Insulin-like growth factor 2 (Igf2) and Insulin, which, while sequence related and very closely linked, bind different receptors. In both cases the receptors are unlinked. Given the sequence relatedness we remove one of the two. In effect we are then asking about a tendency for a ligand cluster to be linked to a receptor cluster. The final data set specifies 108 ligand-receptor sets (106 in the non-augmented set) [see Additional file 1]. Note that most ligands have more than one receptor. When then we refer to ligand receptor "pairs," we refer to incidence in which a ligand is linked to one of its receptors. We also performed the same analyses as given below on a data set in which duplicates are defined exclusively by reference to Blast scores [see Additional file 2]. A list of linked ligands and receptors in this data are provided [see Additional file 3]. Using both data sets, in both augmented and non-augmented form (i.e. with or without the two ligand-receptor pairs highlighted by Copper (1999)) we obtain qualitatively identical results [see Additional file 4].
Mouse data
For analysis of the patterns in mice we identified the orthologs of the human ligands and receptors by reference to the MGI curated set of mouse-human orthologs [54]. For each human gene, the locus link id was cross referenced to the mouse ortholog. Thirteen genes lacking an ortholog were removed. Mouse locus link numbers were employed to access RefSeq numbers, unigene references and the chromosomal locations. Breadth of expression was derived from Unigene cDNA source annotations. The compilation of all mouse genes and their position used in the randomization was derived from MartView [55] at Ensembl requesting those with described LocusLink ids. The positions of the ligands and receptors were found by cross-referencing their LocusLink ids to this Ensembl data set. The few that failed to be resolved by this method were ascribed a position by Blasting their RefSeq against the complete mouse genome [55]. For the randomizations only those genes with well resolved genomic locations were employed (24742 genes).
Randomization and statistics
To ask whether there are more ligand-receptor pairs on the same chromosome than expected by chance, we calculate the observed number and compare this with simulants in which we randomly permute the positions of all ligands and receptors. It is unclear on a priori grounds, however, what should be the null model for the randomization. We consider three possible models.
First we suppose that a ligand or receptor can occur in any location in the genome currently occupied by a protein coding gene. In this instance, in the human genome, the positions permitted in the randomizations correspond to the annotated positions of the 24,300 protein coding genes in the NC_0000n files for the human genome (n from 01–23).
Second, we assume that a ligand or receptor can occur in any location in the genome currently occupied by a protein coding gene with the same or comparable expression breadth. For each of the genes in the complete human and mouse sets we identified the Unigene id by following the LocusLink page pertinent to each gene and identified the breadth of expression in the same way as for the ligand-receptor set. Genes were placed in bins of 0–4, 5–9 tissues etc in which they were expressed. We consider two bin classification systems. In both, ligands and receptors in these randomizations were permitted to reside in the same location as any gene in the same bin from the complete human set.
Third, we suppose that there is something unique about ligands and receptors, such that each ligand or receptor can only be relocated to the position of another ligand or receptor. Breadth of expression is ignored as this too greatly constrains the randomizations.
For each of the above randomization protocols we determined the number of ligand-receptor pairs on the same chromosome. Significance (P) was determined from P = (r+1)/(n+1), where r is the number of simulants with the same or greater number of ligand-receptor pairs than observed in the real data and n is the number of simulants (10,000 in all instances), this being the unbiased estimator [56,57].
To determine whether the ligand-receptor pairs that we observe on a given chromosome are more closely linked than expected we perform two analogous sets of simulations. In the first we calculate the mean distance between these pairs and compare this with the mean of the simulants described above. In the second we ask about the number of ligand-receptor pairs within a given distance of each other (e.g., within 1 Mb). We then compare this number to the mean number found after permuting all genes within the chromosomes within which they are found. By permuting on the same chromosome we control for the number of ligand-receptor pairs on the same chromosome.
All results prove to be insensitive to which of the three randomization null models is employed. Unless stated otherwise a result in the text relates to the first model in the text. All results can be found in attached files, for humans [see Additional file 4] and for mice [see Additional file 5] [see Additional file 6].
Abbreviations
Mb: megabase
Df: degrees of freedom
Authors' contributions
Both authors contributed to analysis, design and write up.
Supplementary Material
Additional file 4
Supplement 4: The results and analysis of randomizations, excluding randomizations controlling for breadth. The results sheet is split in two. The top half ("all genes") refers to analysis of the augmented ligand receptor data set, i.e. the one containing the two ligand-receptor pairs nominated by Cooper. The lower half excludes these two. In each, as detailed in the methods, two filtering methods were used to eliminate tandem duplicates: Blast alone (see Additional file 2) and blast plus close proximity of ligands, close proximity of receptors (see Additional file 1). Within each of these subdivisions two data sets were employed to determine where, in randomizations, a ligand and/or receptor might locate. In one case (lig/rec genes only), the positions of ligands and receptors were interchanged with ligands/receptors derived from the test set. In the second method, possible locations are anywhere in the genome where a gene, of any variety, resides ("all genes in genome"). Randomizations for each of the analysis in turn was performed either permitting a ligand or receptor to locate anywhere in the genome ("all genes") or anywhere on the same chromosome ("within chr"). Analysis was further performed on those ligand receptor sets in which at least one of the genes in the set was placentally expressed ("placental"). N is the number of genes in the test set, "sameChr", refers to the number of incidences of ligand receptor pairs on the same chromosome, "sameChrRand"", is the corresponding number for the randomizations. P_chr is the P value for the comparisons. <d> is the mean distance between ligands and receptors on the same chromosome (in kb) and <dRand> the corresponding number in the randomizations. P_d is the relevant P value. "Count[Nkb]", "countRand, P refer to the number of incidences in the real data set of ligand receptor pairs less than N kb apart, the number in the randomizations and the P value. This statistic is only relevant for the cases where randomization is done within chromosomes. The between chromosome randomization by its nature confounds the effect of the excess of pairs on the same chromosome as well as their proximity.
Click here for file
Additional file 5
Supplement 5: This supplement details the analysis (without control for breadth of expression) in the mouse data sets.
Click here for file
Additional file 6
Supplement 6: This supplement details the analysis with control for breadth of expression in the mouse.
Click here for file
Additional file 7
Supplement 7: This supplement details the evidence for co-duplication for paralogous ligand-receptor pairs
Click here for file
Additional file 1
Supplement 1. The data set of ligand receptor pairs after purging of tandem duplicates by Blast score and by purging of ligands within 1 Mb of each other and receptors within 1 MB of each other. Unigene_id is the Unigene reference number. Entrez_id is the Entrez/LocusLink number. Chr is the chromosome on which the gene is found. Cytog is the cytogenetic location of the gene. Genbank_chr_file is the relevant accession number of the whole chromosome GenBank file. From, to, midpos are the ends of the gene and the middle position of the gene in base pairs, these referring to the locations in the NC files. Number of tissues is a count of the number of entries in cDNA source entry in the Unigene page. Placental: 1 means placentally expressed, 0 means no evidence for placental expression. GI: GI number. RefSeq: the ReqSeq Genbank number
Click here for file
Additional file 2
Supplement 2: The data set of ligand receptor pairs after purging of tandem duplicates by Blast scores alone. Annotation as above.
Click here for file
Additional file 3
Supplement 3: The set of ligand-receptor pairs on the same chromosome in the data set given in supplement 2 (Blast score purged alone). Annotation as for figure 1 in the table. Note: In the above three data sets the final two ligand-receptor pairs are those nominated by Cooper.
Click here for file
Acknowledgements
We thank David N. Cooper for discussion. We thank one referee for helpful comments.
==== Refs
Hurst LD Pal C Lercher MJ The evolutionary dynamics of eukaryotic gene order Nature reviews Genetics 2004 5 299 310 15131653 10.1038/nrg1319
Caron H van Schaik B van der Mee M Baas F Riggins G van Sluis P Hermus MC van Asperen R Boon K Voute PA Heisterkamp S van Kampen A Versteeg R The human transcriptome map: Clustering of highly expressed genes in chromosomal domains Science 2001 291 1289 1292 11181992 10.1126/science.1056794
Lercher MJ Urrutia AO Hurst LD Clustering of housekeeping genes provides a unified model of gene order in the human genome Nature Genetics 2002 31 180 183 11992122 10.1038/ng887
Lercher MJ Urrutia AO Pavlicek A Hurst LD A unification of mosaic structures in the human genome Hum Mol Genet 2003 12 2411 2415 12915446 10.1093/hmg/ddg251
Versteeg R van Schaik BDC van Batenburg MF Roos M Monajemi R Caron H Bussemaker HJ van Kampen AHC The human transcriptome map reveals extremes in gene density, intron length, GC content, and repeat pattern for domains of highly and weakly expressed genes Genome Res 2003 13 1998 2004 12915492 10.1101/gr.1649303
Cohen BA Mitra RD Hughes JD Church GM A computational analysis of whole-genome expression data reveals chromosomal domains of gene expression Nature Genet 2000 26 183 186 11017073 10.1038/79896
Hurst LD Williams EJ Pal C Natural selection promotes the conservation of linkage of co-expressed genes Trends Genet 2002 18 604 606 12446137 10.1016/S0168-9525(02)02813-5
Singer GAC Lloyd AT Huminiecki LB Wolfe KH Clusters of co-expressed genes in mammalian genomes are conserved by natural selection Mol Biol Evol 2005 22 767 775 15574806 10.1093/molbev/msi062
Lercher MJ Blumenthal T Hurst LD Coexpression of neighboring genes in Caenorhabditis elegans is mostly due to operons and duplicate genes Genome Res 2003 13 238 243 12566401 10.1101/gr.553803
Roy PJ Stuart JM Lund J Kim SK Chromosomal clustering of muscle-expressed genes in Caenorhabditis elegans Nature 2002 418 975 979 12214599
Miller MA Cutter AD Yamamoto I Ward S Greenstein D Clustered organization of reproductive genes in the C. elegans genome Curr Biol 2004 14 1284 1290 15268860 10.1016/j.cub.2004.07.025
Spellman PT Rubin GM Evidence for large domains of similarly expressed genes in the Drosophila genome. J Biol 2002 1 5 12144710 10.1186/1475-4924-1-5
Boutanaev AM Kalmykova AI Shevelyou YY Nurminsky DI Large clusters of co-expressed genes in the Drosophila genome Nature 2002 420 666 669 12478293 10.1038/nature01216
Thygesen H Zwinderman A Modelling the correlation between the activities of adjacent genes in drosophila BMC Bioinformatics 2005 6 10 15659243 10.1186/1471-2105-6-10
Williams EJB Bowles DJ Coexpression of neighboring genes in the genome of Arabidopsis thaliana Genome Res 2004 14 1060 1067 15173112 10.1101/gr.2131104
Lee JM Sonnhammer EL Genomic gene clustering analysis of pathways in eukaryotes Genome Res 2003 13 875 882 12695325 10.1101/gr.737703
Wong S Wolfe KH Birth of a metabolic gene cluster in yeast by adaptive gene relocation Nature Genet 2005 37 777 782 15951822 10.1038/ng1584
Fukuoka Y Inaoka H Kohane IS Inter-species differences of co-expression of neighboring genes in eukaryotic genomes BMC Genomics 2004 5 art. no. 4 10.1186/1471-2164-5-4
Teichmann SA Veitia RA Genes encoding subunits of stable complexes are clustered on the yeast chromosomes: An interpretation from a dosage balance perspective Genetics 2004 167 2121 2125 15342545 10.1534/genetics.103.024505
Cooper DN Human Gene Evolution 1999 Oxford, BIOS Scientific
Lennard A Gorman P Carrier M Griffiths S Scotney H Sheer D Solari R Cloning and chromosome mapping of the human interleukin-1 receptor antagonist gene Cytokine 1992 4 83 89 1385987 10.1016/1043-4666(92)90041-O
Wang MH Ronsin C Gesnel MC Coupey L Skeel A Leonard EJ Breathnach R Identification of the Ron gene product as the receptor for the human macrophage stimulating protein Science 1994 266 117 119 7939629
Popovici C Leveugle M Birnbaum D Coulier F Coparalogy: Physical and functional clusterings in the human genome Biochem Biophys Res Commun 2001 288 362 370 11606051 10.1006/bbrc.2001.5794
Haig D Gestational drive and the green-bearded placenta Proc Nat Acad Sci USA 1996 93 6547 6551 8692853 10.1073/pnas.93.13.6547
Peters LL Barker JE Novel inheritance of the murine severe combined anemia and thrombocytopenia (Scat) phenotype. Cell 1993 74 135 142 8334700 10.1016/0092-8674(93)90301-6
Hurst LD scat+ is a selfish gene analogous to Medea of Tribolium castaneum Cell 1993 75 407 408 8221883 10.1016/0092-8674(93)90375-Z
Fisher RA The Genetical Theory of Natural Selection 1930 Oxford, Clarendon Press
Bodmer WF Parsons PA Linkage and recombination in evolution Adv Genet 1962 11 1 100
Nei M Modification of linkage intensity by natural selection Genetics 1967 57 625 641 5583732
Nei M Evolutionary change in linkage intensity Nature 1968 218 1160 1161 5656638
Marquardt T Shirasaki R Ghosh S Andrews SE Carter N Hunter T Pfaff SL Coexpressed EphA receptors and Ephrin-A iigands mediate opposing actions on growth cone navigation from distinct membrane domains Cell 2005 121 127 139 15820684 10.1016/j.cell.2005.01.020
Graeber TG Eisenberg D Bioinformatic identification of potential autocrine signaling loops in cancers from gene expression profiles Nat Genet 2001 29 295 300 11685206 10.1038/ng755
DLRP database
Keightley PD Lercher MJ Eyre-Walker A Evidence for widespread degradation of gene control regions in hominid genomes PLoS Biol 2005 3 282 288 10.1371/journal.pbio.0030042
Keightley PD Eyre-Walker A Deleterious mutations and the evolution of sex Science 2000 290 331 333 11030650 10.1126/science.290.5490.331
Duret L Mouchiroud D Gouy M HOVERGEN - a database of homologous vertebrate genes Nucl Acids Res 1994 22 2360 2365 8036164
McLysaght A Hokamp K Wolfe KH Extensive genomic duplication during early chordate evolution Nat Genet 2002 31 200 204 12032567 10.1038/ng884
Pebusque MJ Coulier F Birnbaum D Pontarotti P Ancient large-scale genome duplications: phylogenetic and linkage analyses shed light on chordate genome evolution Mol Biol Evol 1998 15 1145 1159 9729879
Katsanis N Fitzgibbon J Fisher EMC Paralogy Mapping: Identification of a Region in the Human MHC Triplicated onto Human Chromosomes 1 and 9 Allows the Prediction and Isolation of NovelPBXandNOTCHLoci Genomics 1996 35 101 108 8661110 10.1006/geno.1996.0328
Stankiewicz P Shaw CJ Withers M Inoue K Lupski JR Serial segmental duplications during primate evolution result in complex human genome architecture Genome Res 2004 14 2209 2220 15520286 10.1101/gr.2746604
Veitia RA Exploring the etiology of haploinsufficiency Bioessays 2002 24 175 184 11835282 10.1002/bies.10023
Veitia RA Gene dosage balance: deletions, duplications and dominance Trends Genet 2005 21 33 35 15680512 10.1016/j.tig.2004.11.002
Papp B Pal C Hurst LD Dosage sensitivity and the evolution of gene families in yeast Nature 2003 424 194 197 12853957 10.1038/nature01771
Bourque G Zdobnov EM Bork P Pevzner PA Tesler G Comparative architectures of mammalian and chicken genomes reveal highly variable rates of genomic rearrangements across different lineages Genome Res 2005 15 98 110 15590940 10.1101/gr.3002305
Zhao SY Shetty J Hou LH Delcher A Zhu BL Osoegawa K de Jong P Nierman WC Strausberg RL Fraser CM Human, mouse, and rat genome large-scale rearrangements: Stability versus speciation Genome Res 2004 14 1851 1860 15364903 10.1101/gr.2663304
O'Brien SJ Menotti-Raymond M Murphy WJ Nash WG Wienberg J Stanyon R Copeland NG Jenkins NA Womack JE Graves JAM The promise of comparative genomics in mammals Science 1999 286 458 462 10521336 10.1126/science.286.5439.458
Fares MA Byrne KP Wolfe KH Rate Asymmetry After Genome Duplication Causes Substantial Long Branch Attraction Artifacts in the Phylogeny of Saccharomyces Species Mol Biol Evol 2005 16207937
Popovici C Roubin R Coulier F Birnbaum D An evolutionary history of the FGF superfamily Bioessays 2005 27 849 857 16015590 10.1002/bies.20261
Unigene
Entrez
Steinkasserer A Spurr NK Cox S Jeggo P Sim RB The human IL-1 receptor antagonist gene (IL1RN) maps to chromosome 2q14-q21, in the region of the IL-1 alpha and IL-1 beta loci Genomics 1992 13 654 657 1386337 10.1016/0888-7543(92)90137-H
Dale M Nicklin MJ Interleukin-1 receptor cluster: gene organization of IL1R2, IL1R1, IL1RL2 (IL-1Rrp2), IL1RL1 (T1/ST2), and IL18R1 (IL-1Rrp) on human chromosome 2q Genomics 1999 57 177 179 10191101 10.1006/geno.1999.5767
Wolfe K Evolutionary genomics: yeasts accelerate beyond BLAST Curr Biol 2004 14 R392 394 15186766 10.1016/j.cub.2004.05.015
MGD Human Mouse orthologs
Ensembl MartView
Davison AC Hinkley DV Bootstrap methods and their application 1997 Cambridge, United Kingdom, Cambridge University Press
North BV Curtis D Sham PC A note on the calculation of empirical P values from Monte Carlo procedures Am J Hum Genet 2002 71 439 441 12111669 10.1086/341527
Jin M Chen Y He S Ryan SJ Hinton DR Hepatocyte growth factor and its role in the pathogenesis of retinal detachment Invest Ophthalmol Vis Sci 2004 45 323 329 14691191 10.1167/iovs.03-0355
Schmidt L Duh FM Chen F Kishida T Glenn G Choyke P Scherer SW Zhuang Z Lubensky I Dean M Allikmets R Chidambaram A Bergerheim UR Feltis JT Casadevall C Zamarron A Bernues M Richard S Lips CJ Walther MM Tsui LC Geil L Orcutt ML Stackhouse T Zbar B Germline and somatic mutations in the tyrosine kinase domain of the MET proto-oncogene in papillary renal carcinomas Nat Genet 1997 16 68 73 9140397 10.1038/ng0597-68
Bezerra JA Carrick TL Degen JL Witte D Degen SJ Biological effects of targeted inactivation of hepatocyte growth factor-like protein in mice J Clin Invest 1998 101 1175 1183 9486989
Muraoka RS Sun WY Colbert MC Waltz SE Witte DP Degen JL Friezner Degen SJ The Ron/STK receptor tyrosine kinase is essential for peri-implantation development in the mouse J Clin Invest 1999 103 1277 1285 10225971
Joutel A Corpechot C Ducros A Vahedi K Chabriat H Mouton P Alamowitch S Domenga V Cecillion M Marechal E Maciazek J Vayssiere C Cruaud C Cabanis EA Ruchoux MM Weissenbach J Bach JF Bousser MG Tournier-Lasserve E Notch3 mutations in CADASIL, a hereditary adult-onset condition causing stroke and dementia Nature 1996 383 707 710 8878478 10.1038/383707a0
Miller DL Ortega S Bashayan O Basch R Basilico C Compensation by fibroblast growth factor 1 (FGF1) does not account for the mild phenotypic defects observed in FGF2 null mice Mol Cell Biol 2000 20 2260 2268 10688672 10.1128/MCB.20.6.2260-2268.2000
Kawaguchi H Nakamura K Tabata Y Ikada Y Aoyama I Anzai J Nakamura T Hiyama Y Tamura M Acceleration of Fracture Healing in Nonhuman Primates by Fibroblast Growth Factor-2 J Clin Endocrinol Metab 2001 86 875 880 11158060 10.1210/jc.86.2.875
Weinstein M Xu X Ohyama K Deng CX FGFR-3 and FGFR-4 function cooperatively to direct alveogenesis in the murine lung Development 1998 125 3615 3623 9716527
Zammit C Coope R Gomm JJ Shousha S Johnston CL Coombes RC Fibroblast growth factor 8 is expressed at higher levels in lactating human breast and in breast cancer Br J Cancer 2002 86 1097 1103 11953856 10.1038/sj.bjc.6600213
Muenke M Schell U Hehr A Robin NH Losken HW Schinzel A Pulleyn LJ Rutland P Reardon W Malcolm S A common mutation in the fibroblast growth factor receptor 1 gene in Pfeiffer syndrome Nat Genet 1994 8 269 274 7874169 10.1038/ng1194-269
Xu J Liu Z Ornitz DM Temporal and spatial gradients of Fgf8 and Fgf17 regulate proliferation and differentiation of midline cerebellar structures Development 2000 127 1833 1843 10751172
Hu MC Qiu WR Wang YP Hill D Ring BD Scully S Bolon B DeRose M Luethy R Simonet WS Arakawa T Danilenko DM FGF-18, a novel member of the fibroblast growth factor family, stimulates hepatic and intestinal proliferation Mol Cell Biol 1998 18 6063 6074 9742123
OMIM
|
16277660
|
PMC1309615
|
CC BY
|
2021-01-04 16:29:17
|
no
|
BMC Evol Biol. 2005 Nov 8; 5:62
|
utf-8
|
BMC Evol Biol
| 2,005 |
10.1186/1471-2148-5-62
|
oa_comm
|
==== Front
BMC GenomicsBMC Genomics1471-2164BioMed Central London 1471-2164-6-1521627114010.1186/1471-2164-6-152Research ArticleA genome-wide survey of Major Histocompatibility Complex (MHC) genes and their paralogues in zebrafish Sambrook Jennifer G [email protected] Felipe [email protected] Stephan [email protected] Wellcome Trust Sanger Institute, Genome Campus, Hinxton, Cambridge CB10 ISA, UK2 Max-Planck-Institut für Biologie, Abteilung Immunogenetik, 72076 Tübingen, Germany2005 4 11 2005 6 152 152 5 6 2005 4 11 2005 Copyright © 2005 Sambrook et al; licensee BioMed Central Ltd.2005Sambrook et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The genomic organisation of the Major Histocompatibility Complex (MHC) varies greatly between different vertebrates. In mammals, the classical MHC consists of a large number of linked genes (e.g. greater than 200 in humans) with predominantly immune function. In some birds, it consists of only a small number of linked MHC core genes (e.g. smaller than 20 in chickens) forming a minimal essential MHC and, in fish, the MHC consists of a so far unknown number of genes including non-linked MHC core genes. Here we report a survey of MHC genes and their paralogues in the zebrafish genome.
Results
Using sequence similarity searches against the zebrafish draft genome assembly (Zv4, September 2004), 149 putative MHC gene loci and their paralogues have been identified. Of these, 41 map to chromosome 19 while the remaining loci are spread across essentially all chromosomes. Despite the fragmentation, a set of MHC core genes involved in peptide transport, loading and presentation are still found in a single linkage group.
Conclusion
The results extend the linkage information of MHC core genes on zebrafish chromosome 19 and show the distribution of the remaining MHC genes and their paralogues to be genome-wide. Although based on a draft genome assembly, this survey demonstrates an essentially fragmented MHC in zebrafish.
==== Body
Background
The human Major Histocompatibility Complex (MHC) is a gene-dense region on chromosome 6p21.3, and comprises a group of genes that are involved functionally with the adaptive and innate immune system. From centromere to telomere, it is divided into five regions: extended class II, classical class II, class III, classical class I and extended class I. The classical MHC contains 224 genes, many of which are pseudogenes, and with every two in five expressed genes having a potential immune function and a role in disease resistance [1], this region has become a focal locus in comparative genomics. Antibody and T cell mediated immune responses against invading pathogens are initiated through MHC class I and class II molecules [2]. These main components are not only missing from invertebrates, but are also not present in primitive jawless fish, such as hagfish and lamprey [3]. MHC class I and II molecules do, however, exist in all jawed vertebrates, including the cartilaginous fish. The gene loci in the class III region encode a variety of proteins with both immune and non-immune functions [4].
Genomic sequences encompassing the MHC are currently available for human, chimpanzee, macaque, rat, mouse, cat, pig, horse, quail, chicken, frog, teleost fish and shark [5]. While the genomic architecture of the mammalian MHC is conserved, the number of genes between species can differ greatly. In birds, for example, the minimal essential MHC in chicken consists of 19 genes [6] while gene duplications have expanded this region in quail [7] and sparrows [8]. The linkage of class I, II, and III region genes can be traced back to cartilaginous fish, which are the earliest jawed vertebrates known to have diverged from a common ancestor with humans [9]. MHC loci, however, do not always exist in a single tightly linked cluster as generally observed in mammals. A large scale inversion has separated the class IIB cluster from the MHC-linked class IIA cluster in cattle [10]. The class I and class II loci in zebrafish (Brachydanio rerio) are found on different chromosomes [11,12]. A similar organisation is present in trout, stickleback, common guppy and cichlid fish [13,14]. This observation demonstrates that the separation of the MHC class I and class II loci is characteristic of teleost fish, which represent half of all vertebrates. Since the genes of the immune system were present in the common ancestor of tetrapods and teleosts, the differences in their genomic organisation may be the result of lineage-specific chromosomal events such as duplications, inversions, deletions and translocations.
The genome of the zebrafish is currently being sequenced at the Wellcome Trust Sanger Institute. The availability of sequence data will allow an insight into the understanding and evolution of the immune system in fish. To date, small regions containing major histocompatibility genes have been mapped by radiation hybrid mapping and sequencing of select genomic clones [15,16]. With the ongoing whole-genome sequencing efforts and the compilation of physical maps, it is now possible to examine larger genomic regions and assess the degree of shared synteny between mammals and fish on a genome-wide scale. Here we have utilised a comprehensive whole-genome shotgun assembly data set to fully analyse the zebrafish loci that are related to the human MHC. The human genome contains at least three regions that are paralogous to the MHC [17]. These are thought to be the result of two rounds of duplications that occurred early during vertebrate evolution. In this study, we examine the genome-wide distribution of paralogous genes in zebrafish.
Results and Discussion
Analysis of the MHC class I region in zebrafish
The MHC class I region in human embodies gene clusters coding for HLA class I molecules, histones, solute carriers, vomeronasal receptors, olfactory receptors and zinc fingers [18]. These clusters have undergone a large-scale expansion and each has paralogues throughout the genome. Ten representative members of the extended class I region, which are expressed and have paralogues in the human genome, and more than 30 genes that reside in the classical class I region were chosen for this analysis (Table 1). Genes that were excluded from this study included pseudogenes, zinc fingers with the tripartite motif (TRIM), and RING finger proteins. Approximately half of the human genes were identified in the zebrafish whole-genome assembly, predominantly on chromosome 19. Because multigene families evolve by a birth-and-death process [19], orthologous genes are usually difficult to identify through sequence similarity searches.
Table 1 List of human MHC class I and extended class I (xI) region genes used against the zebrafish whole-genome assembly. Genes with the suffix "like" are either gene fragments and/or highly similar to their human counterparts. Genes not identified by sequence similarity searches are marked as not found (NF). The official gene nomenclature for zebrafish is shown in brackets.
MHC CLASS I REGION GENES
HUMAN ZEBRAFISH
name accession location location location location
xI HFE NM_000410 NF
xI HMGN4 NM_006353 NF
xI PRSS16 NM_005865 NF
xI POM121L2 NM_033482 NF
xI GPX5 NM_001509 NF
xI RFP NM_006510 NF
xI MAS1L NM_052967 NF
xI UBD NM_006398 NF
xI GABBR1 NM_021905 15 [gabbr1] 19 [gabbr1]
xI MOG NM_002433 NF
I HLA* 19 [mhc1uea] 19 [mhc1uaa] 19 [mhc1ufa] 19 [mhc1uda]
s_1723 [mhc1ze] 1 [mhc1-like] (X3) 1 [zgc:64115]
I HCG9 NM_005844 NF
I PPP1R11 NM_021959 NF
I RPP21 NM_024839 NF
I GNL1 NM_005275 16 [gnl1]
I PRR3 NM_025263 NF
I ABCF1 NM_001090 19 [zgc:85667]
I PPP1R10 NM_002714 19 [ppp1r10]
I MRPS18B NM_014046 19 [mrps18b]
I C6orf134 NM_024909 19 [c6orf134]
I C6orf136 NM_145029 19 [c6orf136]
I DHX16 NM_003587 NA5816 [dhx16] 15 [dhx16]
I NRM NM_007243 NA1826 [nrm]
I MDC1 NM_014641 NF
I TUBB NM_178014 20 [tubb5] 14 [tubb2] 23 [tubb2]
I FLOT1 NM_005803 19 [flot1] 16 [flot1b]
I IER3 NM_003897 NF 22 [ier5]
I DDR1 NM_013994 16 [ddr1-like] 4 [ddr2]
I GTF2H4 NM_001517 19 [zgc:77721]
I VARS2L NM_020442 19 [vars2-like]
I DPCR1 NM_080870 NF
I C6orf15 NM_014070 NF
I CDSN NM_001264 NF
I PSORS1C1 NM_014068 NF
I PSORS1C2 NM_014069 NF
I C6orf18 NM_019052 NF
I TCF19 NM_007109 1 [tcf19-like]
I POU5FI NM_002701 16 [pou5f1]
I MICA NM_000247 NF
I HCP5 NM_006674 NF
I MICB NM_005931 NF
Chromosome 19 harbours the largest number of MHC genes in zebrafish (Figure 1). The core region, comprising genes encoding the classical peptide presenting class I molecules (mhc1uea, mhc1uaa, mhc1ufa, mhc1uda), proteasome subunits (psmb8, psmb9a, psmb11, psmb10), ATP-binding cassettes (abcb3, abcb3l1, abcb3l2) and tapasin (tpsn), cluster tightly together and span approximately 175 kb of DNA. Over 5 Mb downstream of these loci are several other genes that span the human MHC region, although these are not all found adjacent to each other. In humans, MHC class I molecules are expressed on the surface of most cells and are recognized by CD8+ cytotoxic T cells. They are heterodimers composed of an alpha subunit, encoded by the MHC, and a beta subunit, or b-2 microglobulin (B2M), which maps to chromosome 15q21–q22.2. In zebrafish too, the b2m gene is found located in a different part of the genome on chromosome 13. Previously unmapped class I ZE lineage genes [20] and novel MHC class I antigen fragments have been identified on chromosome 1 and scaffold 1723. The additional functional class I lineage (mhc1ze) is also present in two other cyprinids (carp and barbus), and not linked to other class I and II genes.
Figure 1 Map of major histocompatibility genes and their paralogues in zebrafish (not to scale). Only chromosomes and two unmapped contigs (NA17761 and NA17767) that harbour MHC-related genes are shown. Orthologues or paralogues of human MHC class I region genes are shown in red, of class II region genes in blue, and class III region genes in green. Genes with the suffix "L" for like are either gene fragments and/or highly similar to their human counterparts. Genes that have more than one copy in the genome are shown by a greater than (>) symbol. Similarities and differences between the whole-genome assembly and mapping information present in the ZFIN database are shown by plus (+) and minus (-) signs, respectively. The core MHC has been compiled from five genomic clones spanning this region [EMBL:AL672216, EMBL:AL672151, EMBL:AL672164, EMBL:AL672176].
Analysis of the MHC class II region in zebrafish
MHC class II molecules are heterodimers comprising A (alpha) and B (beta) chains that present peptides to CD4+ T cells via the endosomal pathway. One class IIA (mhc2daa) and six class IIB (mhc2dab, mhc2dbb, mhc2dcb, mhc2ddb, mhc2deb, mhc2dfb) genes have previously been identified by screening a zebrafish genomic BAC library [21]. Only the mhc2dab and mhc2daa genes are known to be expressed [22]. They are closely linked and were identified on chromosome 8 (Table 2). Analysis of approximately 1 Mb of contiguous DNA surrounding the functional class II region in zebrafish demonstrates the presence of 23 flanking genes mapping to various human chromosomal locations, including two or more genes mapping to human 6q (C6orf117, LOC557721, ppil4), 12q (slc15a4, KIAA1944), 20q (rnpc2, slc12a5), 22q (slc7a4, sf3a1) and Xp (pqbp1l, t541, jarid1c). The lack of genes mapping to the human MHC, in addition to the low gene density of this region, indicates that the functional zebrafish class II region is the result of a translocation event [12].
Table 2 List of human MHC class II and extended class II (xII) region genes used against the zebrafish whole-genome assembly. Genes with the suffix "like" are either gene fragments and/or highly similar to their human counterparts. Genes not identified by sequence similarity searches are marked as not found (NF). The official gene nomenclature for zebrafish is shown in brackets.
MHC CLASS II REGION GENES
HUMAN ZEBRAFISH
name accession location location location location
II C6orf10 NM_006781 NF
II BTNL2 NM_019602 NF
II HLA-D* 8 [mhc2dab] 8 [mhc2daa] 19 [mhc2dcb-rs] s_2399 [mhc2dfb]
5 [mhc2-like] NA14232 [mhc2dbb] NA6696 [mhc2ddb] NA6696 [mhc2-like]
20 [mhc2-like] 15 [mhc2-like]
II TAP2 NM_018833 19 [abcb3] 19 [abcb3l1] 19 [abcb3l2]
II PSMB8 NM_004159 19 [psmb8] 19 [psmb10] 19 [psmb11] 19 [psmb9a]
2 [psmb5] s_1793 [psmb7]
II TAP1 NM_000593 9 [tap1]
II PSMB9 NM_002800 19 [psmb8] 19 [psmb10] 19 [psmb11] 19 [psmb9a]
2 [psmb5] s_1793 [psmb7]
II BRD2 NM_005104 19 [brd2] 9 [brd2-like] s_549 [brd3] 5 [zgc:77289]
14 [brd8] 14 [brdt-like] s_277 [brd4]
xII COL11A2 NM_080679 19 [col11a2] 24 [col11a1]
xII RXRB NM_021976 17 [rxrb] NA16779 [rxrb] 21 [rxra] 5 [rxra]
9 [rxrg]
xII SLC39A7 NM_006979 19 [ke4]
xII HSD17B8 NM_014234 19 [fabgl]
xII RING1 NM_002931 NF 2 [rnf2]
xII VPS52 NM_022553 19 [vps52]
xII RPS18 NM_022551 NA303 [rps18]
xII B3GALT4 NM_003782 NF 22 [zgc:76904] s_1722 [b3galt1]
xII C6orf11 NM_005452 s_2182 [bing4]
xII HKE2 NM_014260 14 [zgc:66282] NA8827 [zgc:66282]
xII RGL2 NM_004761 23 [rgl2] NA14862 [rgl2-like] NA7773 [zgc:77299] NA12682 [ralgds]
xII TAPBP NM_003190 19 [tpsn]
xII ZNF297 NM_005453 19 [znf297]
xII DAXX NM_001350 19 [daxx]
xII LYPLA2L CAB63783 19 [lypla2l-like]
xII KIFC1 NM_002263 19 [kifc1]
Mapping to chromosome 19, approximately 22 Mb telomeric of the class I region, is the class IIB gene mhc2dcb-rs (Q95HJ7), with a similar sequence on contig NA4006. Also within this segment of DNA there are two predicted class IIA chain-encoding gene fragments, consisting of only the alpha2 domain and cytoplasmic tail-encoding parts. It is therefore unlikely that this locus is functional. Assuming that these gene fragments are not due to errors in the whole-genome assembly, it is evidence for a linkage of remnants of MHC-related class II genes with the core MHC region containing the class I peptide presenting genes.
Previously identified mhc2dbb and mhc2dfb genes are predicted to be located in currently unmapped contigs, NA14232 and scaffold 2399. However, mhc2deb (found in clone U08874) was not identified in the current assembly, and may be attributed to gaps in the sequence data or allelic or haplotypic differences. Several fragmented sequences resembling class IIA and/or B chain-encoding genes were also identified in contigs NA17244 and NA15003, and chromosomes 5, 15 and 20. Only contig NA6696 harbours putative complete class IIA and class IIB genes. Another gene found in this 15 kb contig resembles mhc2ddb, consisting of exons 3 and 4 only.
The two TAP transporters in human, namely TAP1 and TAP2, are both located in the MHC class II region, and are closely linked to genes encoding the proteasome subunits and the class II molecules. In zebrafish, however, a TAP1-like sequence is found on chromosomes 9 and three TAP2-like genes are found in the class I region [15] on chromosome 19. The latter, named abcb3, abcb3l1, abcb3l2, form part of the zebrafish core MHC region. The trout tap1 gene is not linked to the major class IA region either [23]. Likewise, in Fugu, the tap1 gene is found on an isolated scaffold that is not linked to the main class I region [24].
Analysis of the MHC class III region in zebrafish
The human class III is the most gene dense region within the MHC containing few or no pseudogenes. Unlike the class I and class II loci that are evolutionary and functionally related [25], the class III region genes are not. The class III, however, include immune-related genes such as those encoding complement components, tumour necrosis factors and heat shock proteins. The search for the MHC class III region in zebrafish was first initiated by the identification of the BF/C2 gene using degenerate PCR [26]. This was then followed by the identification of several, but not all, of the zebrafish homologues of the human MHC class III region genes [16].
In the zebrafish, the MHC class III genes are found distributed throughout their genome (Table 3). Several zebrafish genes (zgc:63773, tnf, bat2-like, ck2b, ddah1, skiv2l, C6orf31-like, ppt2, bat3-like) map to chromosome 19, which also encompasses the largest stretch of MHC-related genes. Two other unmapped scaffolds also harbour several class III genes: neu1, zgc:64108, bat8, zbtb12 are found in NA17767; and bf/c2, rdbp, hsp70, skiv2l are in scaffold NA17761. These are syntenic to the human MHC class III region with the conservation of both gene order and content (Figure 1).
Table 3 List of human MHC class III region genes used against the zebrafish whole-genome assembly. Genes with the suffix "like" are either gene fragments and/or highly similar to their human counterparts. Genes not identified by sequence similarity searches are marked as not found (NF). The official gene nomenclature for zebrafish is shown in brackets.
MHC CLASS III REGION GENES
HUMAN ZEBRAFISH
name accession location location location location
BAT1 NM_004640 19 [zgc:63773] 16 [ddx39]
ATP6V1G2 NM_130463 NF 5 [atp6v1g1]
NFKBIL1 NM_005007 NF
TNF NM_000594 19 [tnf] NA11949 [tnfa-like]
LST1 NM_007161 NF
NCR3 NM_147130 NF
AIF1 NM_001623 NF 5 [zgc:73179]
BAT2 NM_004638 19 [bat2-like]
BAT3 NM_004639 1 [bat3] 19 [bat3-like]
APOM NM_019101 NF
C6orf47 NM_021184 NF
BAT4 NM_033177 s_339 [bat4-like]
CSNK2B NM_001320 19 [ck2b] (X2) 23 [ck2b]
LY6G5B NM_021221 NF
LY6G5C NM_025262 NF
BAT5 NM_021160 NA9087 [bat5-like]
LY6G6D NM_021246 NF
LY6G6E NM_024123 NF
LY6G6C NM_025261 NF
C6orf25 NM_025260 NF
DDAH2 NM_013974 NF 19 [ddah1]
CLIC1 NM_001288 18 [zgc:77538] 14 [zgc:92762] 16 [clica] 17 [clic4]
MSH5 NM_002441 15 [msh5]
C6orf26 NM_025259 NF
C6orf27 NM_025258 1 [c6orf27-like]
VARS2 NM_006295 11 [vars2]
LSM2 NM_021177 6 [smx5]
HSPA1A NM_005345 NA17761 [hsp70] 3 [hsp70] (X3) 8 [hsp70]
C6orf48 NM_016947 NF
NEU1 NM_000434 NA17767 [neu1]
C6orf29 NM_032794 NA17767 [zgc:64108]
BAT8 NM_025256 NA17767 [bat8]
ZBTB12 NM_181842 NA17767 [zbtb12]
BF/C2 NM_001710 21 [bf] NA17761 [bf]
RDBP NM_002904 NA17761 [rdbp]
SKIV2L NM_006929 19 [skiv2l] NA17761 [skiv2l]
DOM3Z NM_032419 NA17766 [dom3z]
STK19 NM_004197 18 [stk19-like]
C4B NM_000592 15 [c4] 15 [a2m] 22 [c3] NA17328 [a2m]
NA14479 [c3]
CYP21A2 NM_000500 6 [cyp21a2]
TNXB NM_032470 16 [tnx-like] 2 [tnw] 2 [tnr] 5 [tnc]
CREBL1 NM_004381 NF 20 [atf6-like]
FKBPL NM_022110 NF
C6orf31 NM_030651 19 [c6orf31-like]
PPT2 NM_005155 19 [ppt2] s_1796 [ppt2] NA5816 [zgc:55621]
EGFL8 NM_030652 NF
AGPAT1 NM_032741 NF 1 [agpat3] 13 [agpat4]
RNF5 NM_006913 NF
AGER NM_001136 NF
PBX2 NM_002586 NA14559 [pbx1a] 11 [pbx1a] 18 [pbx3b] NA11844 [pbxy]
GPSM3 NM_022107 NF
NOTCH4 NM_004557 s_285 [notch3] 5 [notch1] NA15389 [notch2] s_1523 [notch3]
Approximately half of the human MHC class III region genes were not identified in the zebrafish assembly. Among these were the Ly6 family members, which may therefore be mammalian-specific. Alternatively, being involved in the immune response, Ly6 genes evolved more rapidly than others, and might have diverged sufficiently to not be recognised by sequence similarity searches. For eight human genes, ATP6V1G2, AIF1, CLIC1, CREBL1, DDAH2, AGPAT1, PBX2 and NOTCH4, the paralogues but not the orthologues of the genes in the human MHC class III region have been identified by sequence similarity in zebrafish. This observation extends to IER3, RING1 and B3GALT4 found in the class I and extended class II regions. The presence or absence of other genes may be attributed to lineage-specific evolution. For example, lamprey [27], zebrafish [26] and medaka [28] possess genes equally similar to both mammalian BF and C2, while a Xenopus clone has clearly been identified as encompassing BF. This BF/C2 ancestral gene has further duplicated in zebrafish [29], copies of which were identified on chromosome 21 and contig NA17761. Similarly, a recent survey of hsp70 genes in Xiphophorus maculates (platyfish) [30] has revealed that a single HSP70 gene gave rise to four distinct groups of genes: mammalian testis-specific HST70, mammalian MHC-linked HSP70, mammalian HSP70B' and the fish HSP70. Human class III HSPA1A, HSPA1B and HSPA1L genes are intronless and encode identical or near-identical proteins. Intronless zebrafish hsp70 genes were identified on chromosome 8 and scaffold NA17761. Three additional copies of HSP70-like genes were identified on chromosome 3, although these contained one, two or three introns in the 3' end of their sequence. The identified zebrafish genes cluster phylogenetically within the fish subgroup, apart from the sequence on scaffold NA17761, which appears to be similar to the mammalian MHC-linked heat shock protein sequences (data not shown).
Construction of in-silico gene maps
The zebrafish MHC gene map (Figure 1) was constructed using primarily mapping information from the whole-genome assembly displayed in the Ensembl platform. The position of at least 49 genes mapped in this survey could be verified by comparison to experimental data stored in the ZFIN database, which also yielded information for genes residing in unmapped contigs. Mapping data from these two sources did not coincide for several genes, including bat3, clica, vars2, c3, pou5f1, rgl2, tubb5, ddx39, col11a1, pbx1a, and stk19-like. These discrepencies may be attributed to the high levels of polymorphisms and regions of misassembly caused by the source DNA used for the whole-genome assembly being collected from over 1000 embryos. Until the genome sequence is complete, it will not be possible to accurately predict the position and number of all human MHC orthologues in zebrafish. There are also a number of genes, notably CD1, MOG and HFE, that could not be identified in the draft assembly used here (these are listed as 'not found' in Tables 1, 2, 3).
There were a number of difficulties associated with assessing the degree of synteny between the zebrafish whole-genome assembly and the human MHC. Although BLAST is a heavily used analysis tool for identifying related sequences, it does not discriminate between large gene family members. To maximise the identification of MHC-related loci in zebrafish, only the highest BLAST scoring sequences were chosen for further analysis (many of which were unique BLAST reciprocal hits), and searches were conducted using blastp or tblastx. Orthology can be confirmed by obtaining mapping data of surrounding genes, with the assumption that groups of syntenic genes would remain in close proximity through evolution, and therefore be maintained in similar segments of DNA. The physical linkage between numerous MHC-related loci is apparent on zebrafish chromosome 19. Genes mapping to zebrafish chromosome 5 have also been observed to be syntenic with human chromosome 9 [31]. Gene blocks, in particular present in contig NA17761 and NA17767, map closely together in the human MHC class III region. This criterion is more difficult to apply when genes are dispersed, as seen in Figure 1. The Ensembl platform [32] used to map the MHC in zebrafish provides a combination of alignment data, genomic location, detailed transcript structures to compare functional domains of orthologous proteins, in conjunction with multi-species comparisons. In combination with further phylogenetic analyses when duplicates of one gene were identified, the distinction between orthologues and paralogues was ascertained. Nevertheless, it is very difficult to assign orthologous comparative relationships with multigene families until all members of the family have been sequenced in both organisms. This might be the case in human, but the zebrafish genome is yet to be completed.
Duplicated genes and MHC paralagous regions
In human, it was observed that four regions, the MHC on 6p21.3 as well as 9q33–q34, 1q21–q25/1p11–p32 and 19p13.1–p13.4 are paralogous regions [33] that share members of the same gene family. Further analysis has shown that paralogues of human MHC genes are also scattered essentially over all chromosomes [18]. Likewise, genome-wide duplications have recently been examined in zebrafish [31], confirming that an extra round of a whole-genome duplication event occurred early in the teleost lineage after it split from the tetrapod lineage. Evidence that paralogous genes exist in zebrafish became apparent when many duplicates associated with the MHC, and mainly the class III region, were identified (Figure 1). Here we discuss five MHC-encoded gene families in more detail.
The Notch gene family members encode evolutionary transmembrane receptors that regulate cell fate determination. Four Notch paralogues (NOTCH1-4) have been identified in human, and only three of their orthologues are found in zebrafish (Table 3). The orthologue of human NOTCH4, found in the MHC, is absent in the fish whole-genome assembly, indicative that gene loss has followed the block duplication events in teleosts. This is also seen in Fugu [24] and may be due to it being the most divergent member of the Notch family [34]. Two duplicates of notch3 are found in scaffolds 1523 and 285. Until the assembly is complete, it is not clear whether there are two copies of this gene or if this is due to the status of the current assembly. Only one surrounding rdh8-like sequence is found in common between the two scaffolds. In addition, notch1 and notch2 have been mapped to chromosome 5 and contig NA15389.
Retinoid receptors are soluble nuclear proteins belonging to the steroid/thyroid hormone receptor superfamily of transcriptional regulators. The RXR subfamily consists of three polypeptide chains, namely alpha, beta and gamma, encoded by separate loci. The human RXRB gene is found within the MHC, while the RXRA and RXRG paralogues are located on chromosomes 9 and 1, respectively. Five rxr loci have been identified in the zebrafish assembly in contig NA16779 and chromosomes 5, 9, 17 and 21. Two semi-orthologues of human rxrb appear to have arisen from a fish-specific duplication, and duplicates of rxra are also present in the zebrafish genome. Mapping data from the current assembly in comparison to the ZFIN linkage maps are contradictory and may highlight problems in both approaches to mapping genes. Interestingly, neither of the rxrb sequences map to chromosome 19 in the Zv4 whole-genome assembly as originally thought [35].
The human PBX2 gene encodes a homeodomain-containing protein. Three paralogues are located within chromosomes 1q23, 9q33 and 19p12, named PBX1, PBX3 and PBX4, respectively. Four PBX-like sequences were identified in the zebrafish: pbx1a on scaffold NA14559 and chromosome 11; pbx3b on chromosome 18; and pbxy on scaffold NA11844. These are related to human PBX1, PBX3 and PBX4, respectively [36]. The orthologous sequence to human PBX2, which is present in the MHC, has not been identified in the current assembly.
The complement component C4 gene encodes a protein that plays a central role in the innate immune response. Structurally, C4, C5 and C3 belong to the α2 macroglobulin (A2M) protein family and are derived from a single common ancestor. In human, they are found on four different chromosomes: 6p21 (C4), 9q33 (C5), 19p13 (C3) and 12p13 (A2M), and a similar situation exists in all tetrapods studied to date. In zebrafish, the a2m and c4 genes are both found on chromosome 15 as previously shown [37]. Multiple copies of a2m have also been observed in zebrafish with a duplicate being present in NA17328. Two copies of the c3 gene were identified on chromosome 22 and NA14479.
Six members of the chloride intracellular channel (CLIC) gene family (CLIC1–CLIC6) have been described in humans. They are involved in chloride ion transport within various subcellular compartments [38]. Four clic genes were identified in zebrafish on chromosomes 14, 16, 17 and 18. On phylogenetic analysis, they cluster with mammalian CLIC2, CLIC3, CLIC4 and CLIC5 genes, respectively. The orthologue of the human MHC-embedded CLIC1 was not found.
Conclusion
Comparative genomics reveals that the organisation of the MHC in more distantly-related organisms varies from the human model [39]. Teleost fish are particularly unusual in their organisation in comparison to mammals, chicken, Xenopus and shark in that their class I and class II loci are found on different chromosomes [14]. The core class I region in zebrafish, medaka, Fugu and rainbow trout [13,40-42] comprises genes involved in class I peptide presentation and processing: the classical class I molecules, the immunoproteasome subunits, ATP-binding cassette transporters and tapasin. The whole-genome shotgun data for zebrafish has allowed a genome-wide analysis of major histocompatibility genes and their paralogues, and highlights that one or more copies exist of MHC-related genes in fish as in humans. The results obtained thus far extend the linkage information regarding major histocompatibility genes on chromosome 19 in zebrafish and the classical mammalian MHC, and further supports previous findings that the functional class II loci are found on a different chromosome. The distribution of the remaining MHC-related genes and their paralogues is genome-wide, confirming a fragmented MHC in fish.
Methods
The genome of the zebrafish is currently being sequenced at the Sanger Institute [43]. The analysis was carried out on the fourth Ensembl whole-genome assembly Zv4, released on September 2004 [32]. This assembly integrates the whole-genome shotgun assembly with data from the physical map [44]. Protein sequences encoded in 117 human MHC genes were chosen for BLAST sequence similarity searches against the zebrafish Ensembl assembly. Previously identified MHC-related zebrafish cDNAs [45] were also used in the analysis. Gene annotations were verified using VEGA [46]. A number of potentially duplicated genes were identified. Their predicted amino acid sequences were then aligned with other vertebrate gene family members present in the UniProt database [47] using CLUSTAL [48], and the multiple sequence alignments were used for phylogenetic analysis by the neighbour joining method [49] via the PHYLO_WIN interface [50]. Official gene symbols assigned by the Zebrafish Nomenclature Committee (ZNC) and HUGO Nomenclature Committee (HGNC) have been used in the annotation. Zebrafish and human gene names are given in lower and upper case, respectively.
The zebrafish source DNA for the whole-genome assembly was acquired from approximately 1000 five day old embryos. This has resulted in possible misassemblies due to polymorphisms between sequences, different haplotypes, and difficulties with the assembly of duplicated genes and regions, in particular with many highly homologous MHC genes that have arisen by duplication. Where possible, finished clones [e.g. EMBL:AL672216, EMBL:AL672151, EMBL:AL672185, EMBL:AL672164, EMBL:AL672176] were used to study the short-range linkage of genes in preference to the Ensembl whole-genome assembly. Alternative mapping information, for comparison and validity of the in-silico approach used during this study, was obtained from the Zebrafish Information Network (ZFIN) [45] and published scientific literature.
Authors' contributions
JS carried out the analysis on the whole-genome assembly and drafted the manuscript. FF contributed materials and data. SB conceived the study and contributed towards the preparation of the manuscript.
Note added in proof
At least one of the genes stated to be missing in assembly Zv4 is present in assembly Zv5 which was released on October 2005.
Acknowledgements
We like to thank all members of the sequencing department and zebrafish project group at the Sanger Institute [43] and K. Belov for comments. We would like to acknowledge J. Klein for providing genomic clones used in this study. JS and SB were supported by the Wellcome Trust.
==== Refs
MHC Sequencing Consortium Complete sequence and gene map of a human major histocompatibility complex. The MHC sequencing consortium Nature 1999 401 921 923 10553908 10.1038/44853
Parham P Virtual reality in the MHC Immunol Rev 1999 167 5 15 10319247
Marchalonis JJ Schluter SF On the relevance of invertebrate recognition and defence mechanisms to the emergence of the immune response of vertebrates Scand J Immunol 1990 32 13 20 2382105
Milner CM Campbell RD Genetic organization of the human MHC class III region Front Biosci 2001 6 D914 26 11487476
Kelley J Walter L Trowsdale J Comparative genomics of major histocompatibility complexes Immunogenetics 2005 56 683 695 15605248 10.1007/s00251-004-0717-7
Kaufman J Milne S Gobel TW Walker BA Jacob JP Auffray C Zoorob R Beck S The chicken B locus is a minimal essential major histocompatibility complex Nature 1999 401 923 925 10553909 10.1038/44856
Shiina T Shimizu S Hosomichi K Kohara S Watanabe S Hanzawa K Beck S Kulski JK Inoko H Comparative genomic analysis of two avian (quail and chicken) MHC regions J Immunol 2004 172 6751 6763 15153492
Freeman-Gallant CR Johnson EM Saponara F Stanger M Variation at the major histocompatibility complex in Savannah sparrows Mol Ecol 2002 11 1125 1130 12030987 10.1046/j.1365-294X.2002.01508.x
Terado T Okamura K Ohta Y Shin DH Smith SL Hashimoto K Takemoto T Nonaka MI Kimura H Flajnik MF Nonaka M Molecular cloning of C4 gene and identification of the class III complement region in the shark MHC J Immunol 2003 171 2461 2466 12928394
Band M Larson JH Womack JE Lewin HA A radiation hybrid map of BTA23: identification of a chromosomal rearrangement leading to separation of the cattle MHC class II subregions Genomics 1998 53 269 275 9799592 10.1006/geno.1998.5507
Bingulac-Popovic J Figueroa F Sato A Talbot WS Johnson SL Gates M Postlethwait JH Klein J Mapping of mhc class I and class II regions to different linkage groups in the zebrafish, Danio rerio Immunogenetics 1997 46 129 134 9162099 10.1007/s002510050251
Kuroda N Figueroa F O'HUigin C Klein J Evidence that the separation of Mhc class II from class I loci in the zebrafish, Danio rerio, occurred by translocation Immunogenetics 2002 54 418 430 12242592 10.1007/s00251-002-0473-5
Hansen JD Strassburger P Thorgaard GH Young WP Du Pasquier L Expression, linkage, and polymorphism of MHC-related genes in rainbow trout, Oncorhynchus mykiss J Immunol 1999 163 774 786 10395670
Sato A Figueroa F Murray BW Malaga-Trillo E Zaleska-Rutczynska Z Sultmann H Toyosawa S Wedekind C Steck N Klein J Nonlinkage of major histocompatibility complex class I and class II loci in bony fishes Immunogenetics 2000 51 108 116 10663573 10.1007/s002510050019
Michalova V Murray BW Sultmann H Klein J A contig map of the Mhc class I genomic region in the zebrafish reveals ancient synteny J Immunol 2000 164 5296 5305 10799891
Sultmann H Sato A Murray BW Takezaki N Geisler R Rauch GJ Klein J Conservation of Mhc class III region synteny between zebrafish and human as determined by radiation hybrid mapping J Immunol 2000 165 6984 6993 11120825
Kasahara M Genome dynamics of the major histocompatibility complex: insights from genome paralogy Immunogenetics 1999 50 134 145 10602875 10.1007/s002510050589
Horton R Wilming L Rand V Lovering RC Bruford EA Khodiyar VK Lush MJ Povey S Talbot CCJ Wright MW Wain HM Trowsdale J Ziegler A Beck S Gene map of the extended human MHC Nat Rev Genet 2004 5 889 899 15573121 10.1038/nrg1489
Nei M Gu X Sitnikova T Evolution by the birth-and-death process in multigene families of the vertebrate immune system Proc Natl Acad Sci U S A 1997 94 7799 7806 9223266 10.1073/pnas.94.15.7799
Kruiswijk CP Hermsen TT Westphal AH Savelkoul HF Stet RJ A novel functional class I lineage in zebrafish (Danio rerio), carp (Cyprinus carpio), and large barbus (Barbus intermedius) showing an unusual conservation of the peptide binding domains J Immunol 2002 169 1936 1947 12165519
Graser R Vincek V Takami K Klein J Analysis of zebrafish Mhc using BAC clones Immunogenetics 1998 47 318 325 9472068 10.1007/s002510050364
Sultmann H Mayer WE Figueroa F O'Huigin C Klein J Organization of Mhc class II B genes in the zebrafish (Brachydanio rerio) Genomics 1994 23 1 14 7829056 10.1006/geno.1994.1452
Phillips RB Zimmerman A Noakes MA Palti Y Morasch MR Eiben L Ristow SS Thorgaard GH Hansen JD Physical and genetic mapping of the rainbow trout major histocompatibility regions: evidence for duplication of the class I region Immunogenetics 2003 55 561 569 14566436 10.1007/s00251-003-0615-4
Sambrook JG Russell R Umrania Y Edwards YJ Campbell RD Elgar G Clark MS Fugu orthologues of human major histocompatibility complex genes: a genome survey Immunogenetics 2002 54 367 380 12242588 10.1007/s00251-002-0478-0
Klein J Sato A Birth of the major histocompatibility complex Scand J Immunol 1998 47 199 209 9519857 10.1046/j.1365-3083.1998.00292.x
Seeger A Mayer WE Klein J A complement factor B-like cDNA clone from the zebrafish (Brachydanio rerio) Mol Immunol 1996 33 511 520 8700167 10.1016/0161-5890(96)00002-8
Nonaka M Takahashi M Sasaki M Molecular cloning of a lamprey homologue of the mammalian MHC class III gene, complement factor B J Immunol 1994 152 2263 2269 7510741
Kuroda N Wada H Naruse K Simada A Shima A Sasaki M Nonaka M Molecular cloning and linkage analysis of the Japanese medaka fish complement Bf/C2 gene Immunogenetics 1996 44 459 467 8824158 10.1007/s002510050153
Gongora R Figueroa F Klein J Independent duplications of Bf and C3 complement genes in the zebrafish Scand J Immunol 1998 48 651 658 9874500 10.1046/j.1365-3083.1998.00457.x
Yamashita M Hirayoshi K Nagata K Characterization of multiple members of the HSP70 family in platyfish culture cells: molecular evolution of stress protein HSP70 in vertebrates Gene 2004 336 207 218 15246532 10.1016/j.gene.2004.04.023
Woods IG Wilson C Friedlander B Chang P Reyes DK Nix R Kelly PD Chu F Postlethwait JH Talbot WS The zebrafish gene map defines ancestral vertebrate chromosomes Genome Res 2005 15 1307 1314 16109975 10.1101/gr.4134305
Ensembl
Flajnik MF Kasahara M Comparative genomics of the MHC: glimpses into the evolution of the adaptive immune system Immunity 2001 15 351 362 11567626 10.1016/S1074-7613(01)00198-4
Li L Huang GM Banta AB Deng Y Smith T Dong P Friedman C Chen L Trask BJ Spies T Rowen L Hood L Cloning, characterization, and the complete 56.8-kilobase DNA sequence of the human NOTCH4 gene Genomics 1998 51 45 58 9693032 10.1006/geno.1998.5330
Gongora R Zaleska-Rutczynska Z Takami K Figueroa F Klein J Linkage of RXRB-like genes to class I and not to class II Mhc genes in the zebrafish Immunogenetics 1998 48 141 143 9634478 10.1007/s002510050414
Abi-Rached L Gilles A Shiina T Pontarotti P Inoko H Evidence of en bloc duplication in vertebrate genomes Nat Genet 2002 31 100 105 11967531 10.1038/ng855
Samonte IE Sato A Mayer WE Shintani S Klein J Linkage relationships of genes coding for alpha2-macroglobulin, C3 and C4 in the zebrafish: implications for the evolution of the complement and Mhc systems Scand J Immunol 2002 56 344 352 12234255 10.1046/j.1365-3083.2002.01154.x
Jentsch TJ Stein V Weinreich F Zdebik AA Molecular structure and physiological function of chloride channels Physiol Rev 2002 82 503 568 11917096
Kulski JK Shiina T Anzai T Kohara S Inoko H Comparative genomic analysis of the MHC: the evolution of class I duplication blocks, diversity and complexity from shark to man Immunol Rev 2002 190 95 122 12493009 10.1034/j.1600-065X.2002.19008.x
Takami K Zaleska-Rutczynska Z Figueroa F Klein J Linkage of LMP, TAP, and RING3 with Mhc class I rather than class II genes in the zebrafish J Immunol 1997 159 6052 6060 9550404
Clark MS Shaw L Kelly A Snell P Elgar G Characterization of the MHC class I region of the Japanese pufferfish (Fugu rubripes) Immunogenetics 2001 52 174 185 11220619 10.1007/s002510000285
Matsuo MY Asakawa S Shimizu N Kimura H Nonaka M Nucleotide sequence of the MHC class I genomic region of a teleost, the medaka (Oryzias latipes) Immunogenetics 2002 53 930 940 11862394 10.1007/s00251-001-0427-3
The Danio rerio Sequencing Project
Zebrafish Genome Fingerprinting Project
Sprague J Clements D Conlin T Edwards P Frazer K Schaper K Segerdell E Song P Sprunger B Westerfield M The Zebrafish Information Network (ZFIN): the zebrafish model organism database Nucleic Acids Res 2003 31 241 243 12519991 10.1093/nar/gkg027
VEGA
Bairoch A Apweiler R Wu CH Barker WC Boeckmann B Ferro S Gasteiger E Huang H Lopez R Magrane M Martin MJ Natale DA O'Donovan C Redaschi N Yeh LS The Universal Protein Resource (UniProt) Nucleic Acids Res 2005 33 D154 9 15608167 10.1093/nar/gki070
Higgins DG Thompson JD Gibson TJ Using CLUSTAL for multiple sequence alignments Methods Enzymol 1996 266 383 402 8743695
Saitou N Nei M The neighbor-joining method: a new method for reconstructing phylogenetic trees Mol Biol Evol 1987 4 406 425 3447015
Galtier N Gouy M Gautier C SEAVIEW and PHYLO_WIN: two graphic tools for sequence alignment and molecular phylogeny Comput Appl Biosci 1996 12 543 548 9021275
|
16271140
|
PMC1309616
|
CC BY
|
2021-01-04 16:32:47
|
no
|
BMC Genomics. 2005 Nov 4; 6:152
|
utf-8
|
BMC Genomics
| 2,005 |
10.1186/1471-2164-6-152
|
oa_comm
|
==== Front
BMC Infect DisBMC Infectious Diseases1471-2334BioMed Central London 1471-2334-5-1011627115710.1186/1471-2334-5-101Research ArticleAdjunctive corticosteroids for Pneumocystis jiroveci pneumonia in patients with HIV infection: a meta-analysis of randomised controlled trials Briel Matthias [email protected] Remy [email protected] Hansjakob [email protected] Heiner C [email protected] Institut für klinische Epidemiologie, Universitätsspital Basel, 4031 Basel, Switzerland2 Klinik und Poliklinik für Infektiologie, Inselspital, 3010 Bern, Switzerland3 Klinik für Infektiologie, Universitätsspital Basel, 4031 Basel, Switzerland2005 7 11 2005 5 101 101 27 6 2005 7 11 2005 Copyright © 2005 Briel et al; licensee BioMed Central Ltd.2005Briel et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The objective of this study was to review the effects of adjunctive corticosteroids on overall mortality and the need for mechanical ventilation in HIV-infected patients with Pneumocystis jiroveci pneumonia (PCP) and substantial hypoxemia (arterial oxygen partial pressure <70 mmHg or alveolar-arterial gradient >35 mmHg on room air).
Methods
We conducted a systematic search of the literature for randomised trials published up to December 2004. Selected trials compared adjunctive corticosteroids with placebo or usual care in HIV-infected patients with PCP and reported mortality data. Two teams of reviewers independently evaluated the methodology and extracted data from each primary study.
Results
Six studies were included in the meta-analysis. Risk ratios for overall mortality for adjunctive corticosteroids were 0.54 (95% confidence interval [CI], 0.38–0.79) at 1 month and 0.67 (95% CI, 0.49–0.93) at 3–4 months of follow-up. Numbers needed to treat, to prevent 1 death, are 9 patients in a setting without highly active antiretroviral therapy (HAART) available and 22 patients with HAART available. Only the 3 largest trials provided data on the need for mechanical ventilation with a risk ratio of 0.37 (95% CI, 0.20–0.70) in favour of adjunctive corticosteroids.
Conclusion
The number and size of trials investigating adjunctive corticosteroids for HIV-infected patients with PCP is small, but our results suggest a beneficial effect for patients with substantial hypoxemia.
==== Body
Background
With the introduction of highly active antiretroviral therapy (HAART) more than a decade ago, the incidence of Pneumocystis jiroveci pneumonia (PCP) [1] has decreased significantly in the Western hemisphere. However, PCP still remains the most common opportunistic infection in patients infected with the human immunodeficiency virus (HIV) [2]. Among patients with HIV infection and PCP the mortality rate is 10 to 20% during the initial infection and increases substantially with the need for mechanical ventilation [3]. In 1990 an expert panel recommended the use of corticosteroids for HIV-infected patients with PCP and substantial hypoxemia (initial arterial oxygen partial pressure of <70 mmHg or alveolar-arterial gradient >35 mmHg on room air) based on the evidence from five randomised controlled trials [4]. This consensus statement still represents the basis of current treatment guidelines [5]. However, at the time this statement was made, one trial was not yet completed [6], two trials were stopped prematurely [7,8], and one trial was not published in full [9]. In 1992 a systematic review qualitatively summarised the same incomplete data [10].
We present an updated systematic review and meta-analysis of randomised controlled trials to assess the magnitude of effects of adjunctive corticosteroid therapy on overall mortality and the need for mechanical ventilation in HIV-related PCP. In addition, we provide numbers needed to treat that may serve as estimates for the expected benefit of adjunctive corticosteroid therapy in the presence and absence of HAART.
Methods
Search for relevant studies
We searched MEDLINE (January 1985 – December 2004), EMBASE (January 1985 – December 2004) and the Cochrane Library (issue 4, 2004) without language restrictions to identify randomised controlled trials that compared adjunctive corticosteroids to control in HIV-infected patients with PCP. We used the terms steroid*, corticosteroid*, glucocorticoid*, pneumocystis, PCP, *carinii, *jiroveci as text words and Glucocorticoids, Adrenal Cortex Hormones, Steroids, Pneumocystis Infections, Pneumocystis jiroveci, and Pneumonia, Pneumocystis as Medical Subject Headings. We restricted the search to articles indexed as randomised controlled trials (publication type) or drug therapy (subject heading) or those that included the words random* or placebo in their titles or abstracts. We further reviewed the reference lists from previously published overviews [4,10], we searched UptoDate version 2005 and Clinical Evidence Concise (issue 12, 2004), contacted experts of the field, and searched reference lists of identified publications for citations of additional relevant articles.
Study selection and data abstraction
Trials were considered eligible for this meta-analysis if they compared corticosteroids to placebo or usual care in HIV-infected patients with PCP in addition to baseline treatment with trimethoprim-sulfamethoxazole, pentamidine or dapsone-trimethoprim, used random allocation, and reported mortality data. We excluded trials in patients with no or mild hypoxemia (arterial oxygen partial pressure >70 mmHg or an alveolar-arterial gradient <35 mmHg on room air) and trials with a follow-up of less than 30 days.
Two teams of investigators (MB/HCB and RB/HF) assessed study eligibility and quality blinded to one another's rating and resolved any disagreement by consensus. Data of eligible trials were abstracted in duplicate. We assessed the quality of included trials with respect to concealment of treatment allocation; blinding of patients, caregivers or assessors of clinical outcomes; performance of a sample size calculation; and if the trial was stopped early [11]. The main endpoint for benefit of adjunctive corticosteroid therapy was overall mortality. A secondary endpoint was the need for mechanical ventilation.
Statistical analysis
All analysis was according to the intent-to-treat principle. We pooled treatment effects across studies and calculated a weighted average risk ratio of overall mortality in the treatment and control groups by using a random effects model. We investigated the presence of publication bias by means of funnel plots [12]. We tested for heterogeneity with the Cochran Q test and measured inconsistency (I2; the percentage of total variance across studies that is due to heterogeneity rather than chance) of treatment effects across studies [13,14]. We carried out sensitivity analyses to examine treatment effects according to quality components of trials, and if publication was as a peer-reviewed article or just in abstract form. Numbers needed to treat were calculated by multiplying the mean relative risk reduction with an initial mean baseline risk [15]. All statistical analyses were performed using Stata 8.2 (StataCorp, College Station, Tex).
Results
We identified 8 trials [6-9,16-19] that met our inclusion criteria (Figure 1). We excluded one trial [17] because it investigated only patients with mild hypoxemia and had a very short follow-up of only 3 days, and another trial [18] which turned out to be a subgroup analysis of a larger included trial [8] [see Additional file 1]. In total, there were 242 individuals in the intervention and 247 individuals in the control groups. The Funnel plot indicated no evidence for a publication bias (Figure 2). Characteristics of included trials are provided in Table 1.
Figure 1 Flow diagram of trials. RCT, randomised controlled trial.
Figure 2 Funnel plot to evaluate the presence of a publication bias in trials investigating adjunctive corticosteroids for pneumocystis jiroveci pneumonia in HIV-infected patients. The funnel graph plots the log of the treatment odds ratio against the standard error (s.e.) of the log odds ratio (an indicator of sample size). Open circles represent trials included in the meta-analysis. The line in the centre indicates the summary log odds ratio. In the absence of a publication bias, the log odds ratio estimates from smaller trials are expected to be scattered above and below the summary estimate, producing a symmetric triangular or funnel shape. When smaller trials with larger log odds ratios are missing, the funnel plot appears asymmetric and may indicate the presence of a publication bias. In our systematic review the funnel plot looks symmetric. The Egger test for publication bias was not statistically significant (P = 0.91).
Table 1 Characteristics of included trials
Author (year) reference Diagnosis of PCP Baseline treatment for PCP Oxygenation entry criteria Corticoid (route)/initial daily dose/duration (days) Interval (max.) * Blinding patients/care givers/assessors Concealed allocation/n centres Sample size calculation/stopped early Randomised individuals to I/C Total deaths in I/C
Clement et al. (1989) [9] BAL, sputum 88% TMP-SMX, 12% Pentamidine PaO2<51 mmHg (room air) Methylprednisolon (IV)/240 mg/8 d Unlimited Yes/Yes/Unclear Unclear/1 Unclear/No 19/22 9/9 at 56 days
Montaner et al. (1990) [8] BAL TMP-SMX, Pentamidine, Dapsone-TMP 85–90% O2-Saturation † Prednisone (oral)/60 mg/7 d with 14 d tapering 48 h Yes/Yes/Unclear Yes/1 Yes/Yes 18/19 1/0 at 30 days 2/1 at 90 days
Bozzette et al. (1990) [16] 75% BAL, 15% sputum + presumed 80% TMP-SMX, 18% Pentamidine, 2% Dapsone-TMP Hypoxemia ratio >75 ‡ Prednisone (oral)/80 mg/21 d or as baseline treatment 36 h No/No/Unclear Yes/6 Yes/No 123/128 13/28 at 31 days 20/33 at 84 days
Gagnon et al. (1990) [7] BAL, biopsy, sputum TMP-SMX PaO2<75 mmHg (35% oxygen) Methylprednisolon (IV)/160 mg/7–10 d 72 h Yes/Yes/Unclear Unclear/1 Yes/Yes 12/11 3/9 at 28 days 5/9 at 120 days
Nielsen et al. (1992) [6] BAL, biopsy TMP-SMX PaO2<67.5 mmHg (room air) Methylprednisolon (IV)/2 mg/kg/10 d 24 h No/No/Unclear Unclear/3 Unclear/Yes 30/29 2/9 at 34 days 4/9 at 90 days
Walmsley et al. (1995) [19] BAL, biopsy, sputum 82% TMP-SMX, 17% Pentamidine, 1% Dapsone-TMP PaO2<70 mmHg (room air) § Methylprednisolon (IV)/80 mg/10 d 24 h Yes/Yes/Unclear Yes/3 Yes/No 40/38 4/6 at 35 days
Abbreviations: PCP, Pneumocystis jiroveci pneumonia; n, number; I/C, intervention/control group; BAL, bronchoalveolar lavage; TMP, trimethoprim; SMX, sulfamethoxazole; IV, intravenous.
* Maximal interval between initiation of baseline treatment for Pneumocystis jiroveci pneumonia and initiation of corticosteroid.
† Or 5% decrease on exercise.
‡ PaO2 divided by fraction of inspired oxygen.
§ Or alveolar-arterial oxygen gradient >40 mmHg if arterial blood gases could not be assessed on room air.
Overall mortality
Risk ratios for overall mortality were significantly reduced for adjunctive corticoids at 1 month (0.54;95% CI, 0.38–0.79) and at 3–4 months (0.67;95% CI, 0.49–0.93) of follow-up (Figure 3). We found some evidence for heterogeneity among trials at 1 month (test of heterogeneity, P = 0.12; I2 = 43% [95% uncertainty interval [UI], 0%–78%]) whereas at 3–4 months treatment effects looked more homogenous (P = 0.46; I2 = 0% [95% UI, 0%–75%]. In a sensitivity analysis heterogeneity was considerably reduced when the analysis was limited to trials with early (<3 days) adjunctive corticosteroids that were published in full, i.e. excluding Clement et al. [9] (summary risk ratio for mortality at 1 month: 0.45 (95% CI, 0.29–0.70), heterogeneity P = 0.49; I2 = 0% [95% UI, 0%–79%]). In further sensitivity analyses for the mortality endpoint at 1 month summary risk ratios were 0.55 (95% CI, 0.32–0.93) in trials that reported concealed allocation [8,16,19], 0.74(95% CI, 0.45–1.21) in trials reporting blinding of patients and care-givers [7-9,19], and 0.64 (95% CI, 0.42–0.98) in trials not prematurely halted [9,16,19].
Figure 3 Summary estimates for overall mortality at 1 month (A) and 3–4 months (B) follow-up. The Cochran Q test for heterogeneity. I2 as a measure of inconsistency (in percent). CI indicates confidence interval; UI, uncertainty interval.
Need for mechanical ventilation
Reliable data on the need for mechanical ventilation was only available for the 3 largest trials [6,16,19]. Again, the risk ratio for this endpoint was largely reduced in the group with early adjunctive corticosteroids (0.37;95% CI, 0.20–0.70; P = 0.40; I2 = 0% [95% UI, 0%–90%]).
Discussion
This systematic review of 6 randomised controlled trials in HIV-infected patients with PCP and substantial hypoxemia found a significant relative risk reduction of death for adjunctive corticosteroids of 46% at 1 month and of 33% at3–4 months. The average-weighted mean mortality in control groups of included trials at one month was 25%. This initial mortality-rate of 25% can be assumed in settings where HAART is not available which is still the case for most developing countries [20]. In this situation we estimated that 9 (95% CI, 6–19) HIV-infected patients with PCP have to be treated early with adjunctive corticosteroids to prevent 1 death during the first month after PCP diagnosis. In Western countries, where HAART is widely available, the respective number to treat was estimated to be 22 (95% CI, 16–48) patients assuming an initial mortality rate of 10% [21]. With regard to the need for mechanical ventilation the risk reduction for adjunctive corticosteroids was even greater in the investigated patient population, but the number of trials was small (n = 3).
Our study has several strengths and limitations. We conducted an extensive literature search to retrieve all eligible trials. However, formal testing for publication bias was not very powerful because of a relatively small number of included trials. Even with a symmetric looking funnel plot, such bias cannot be ruled out. Moreover, with a small number of included trials the uncertainty interval for the inconsistency among trials may not be very informative [13]. We focused mainly on mortality data that may be less prone to ascertainment bias, and we analysed the data according to the intent-to-treat principle to get more conservative estimates. Finally, the trials included in this meta-analysis used different corticosteroid regimen. So far, neither the dosing nor the length and tapering schedule of corticosteroids has been addressed adequately in randomised trials. In current recommendations [5] the corticosteroid schedule of the largest trial [16] was adapted.
There has been some concern among physicians treating patients with AIDS that further immunosuppression due to corticosteroid therapy could accelerate the onset of other HIV-related opportunistic complications [22,23]. However, with the exception of an increase in muco-cutaneous herpes simplex infection episodes [16], adjunctive corticosteroids were not associated with an increase in opportunistic complications in any of the included trials. A large cohort study which used a standard 21-day tapering course of adjunctive corticosteroids found no difference in the risk of AIDS-related complications apart from an increase in esophageal candidiasis [24].
It is possible that adjunctive corticosteroids are also beneficial for HIV-infected patients with mild hypoxemia due to PCP [17]. However, in this situation the short term mortality is low and possible unfavourable effects of corticosteroids might outweigh the benefits. Moreover, corticosteroids might also be beneficial for non-HIV-infected patients with severe PCP [25], but evidence from randomised controlled trials is still lacking.
Conclusion
This meta-analysis confirmed and quantified the benefit of adjunctive corticosteroid therapy in HIV-infected patients with moderate-severe PCP. We estimated a relative risk reduction for overall mortality of 46% at 1 month and 33% at 3–4 months. We calculated numbers needed to treat of 9 patients for settings without HAART, and 22 patients with HAART available to prevent 1 death. Our results underline the conclusions of the early released consensus statement [4], and support current recommendations for the management of PCP in HIV-infected patients [5].
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
MB and HCB conceived of the study and performed the literature search. MB, HCB, RB and HF checked eligibility and quality of trials, and extracted the necessary data. MB performed the statistical analyses and drafted the manuscript with the help of HCB, RB and HF. All authors read and approved the final version.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Supplementary Material
Additional File 1
Table with characteristics of excluded trials and reasons for exclusion.
Click here for file
Acknowledgements
We would like to thank M. Simcock for proofreading the manuscript.
There was no external funding for the study. H.C. Bucher and M. Briel are supported by Santésuisse and the Gottfried and Julia Bangerter-Rhyner Foundation, Switzerland. These institutions had no role in the study design, data collection, data analysis, data interpretation, writing and submission of the manuscript for publication.
==== Refs
Stringer JR Beard CB Miller RF Wakefield AE A new name (Pneumocystis jiroveci) for Pneumocystis from humans Emerg Infect Dis 2002 8 891 896 12194762
Kaplan JE Hanson D Dworkin MS Frederick T Bertolli J Lindegren ML Holmberg S Jones JL Epidemiology of human immunodeficiency virus-associated opportunistic infections in the United States in the era of highly active antiretroviral therapy Clin Infect Dis 2000 30 Suppl 1 S5 14 10770911 10.1086/313843
Randall CJ Yarnold PR Schwartz DN Weinstein RA Bennett CL Improvements in outcomes of acute respiratory failure for patients with human immunodeficiency virus-related Pneumocystis carinii pneumonia Am J Respir Crit Care Med 2000 162 393 398 10934059
Consensus statement on the use of corticosteroids as adjunctive therapy for pneumocystis pneumonia in the acquired immunodeficiency syndrome. The National Institutes of Health-University of California Expert Panel for Corticosteroids as Adjunctive Therapy for Pneumocystis Pneumonia N Engl J Med 1990 323 1500 1504 2136587
Benson CA Kaplan JE Masur H Pau A Holmes KK Treating opportunistic infections among HIV-exposed and infected children: recommendations from CDC, the National Institutes of Health, and the Infectious Diseases Society of America MMWR Recomm Rep 2004 53 1 112 15841069
Nielsen TL Eeftinck Schattenkerk JK Jensen BN Lundgren JD Gerstoft J van Steenwijk RP Bentsen K Frissen PH Gaub J Orholm M Hansen J Mathiesen L Skinhoj P Danner SA Nielsen JO Adjunctive corticosteroid therapy for Pneumocystis carinii pneumonia in AIDS: a randomized European multicenter open label study J Acquir Immune Defic Syndr 1992 5 726 731 1613673
Gagnon S Boota AM Fischl MA Baier H Kirksey OW La Voie L Corticosteroids as adjunctive therapy for severe Pneumocystis carinii pneumonia in the acquired immunodeficiency syndrome. A double-blind, placebo-controlled trial N Engl J Med 1990 323 1444 1450 2233916
Montaner JS Lawson LM Levitt N Belzberg A Schechter MT Ruedy J Corticosteroids prevent early deterioration in patients with moderately severe Pneumocystis carinii pneumonia and the acquired immunodeficiency syndrome (AIDS) Ann Intern Med 1990 113 14 20 2190515
Clement M Edison R Turner J Montgomery B Luce J Feigal D Hopewell P Corticosteroids as adjunctive therapy in severe Pneumocystis carinii pneumonia: a prospective placebo-controlled trial. Am Rev Respir Dis 1989 139 A250 (abstract)
Sistek CJ Wordell CJ Hauptman SP Adjuvant corticosteroid therapy for Pneumocystis carinii pneumonia in AIDS patients Ann Pharmacother 1992 26 1127 1133 1421680
Juni P Witschi A Bloch R Egger M The hazards of scoring the quality of clinical trials for meta-analysis JAMA 1999 282 1054 1060 10493204 10.1001/jama.282.11.1054
Sterne JA Egger M Smith GD Systematic reviews in health care: Investigating and dealing with publication and other biases in meta-analysis BMJ 2001 323 101 105 11451790 10.1136/bmj.323.7304.101
Higgins JP Thompson SG Deeks JJ Altman DG Measuring inconsistency in meta-analyses BMJ 2003 327 557 560 12958120 10.1136/bmj.327.7414.557
Higgins JP Thompson SG Quantifying heterogeneity in a meta-analysis Stat Med 2002 21 1539 1558 12111919 10.1002/sim.1186
Marx A Bucher HC Numbers needed to treat derived from meta-analysis: a word of caution ACP J Club 2003 138 A11 A12 12614141
Bozzette SA Sattler FR Chiu J Wu AW Gluckstein D Kemper C Bartok A Niosi J Abramson I Coffman J Hughlett C Loya R Cassens B Akil B Meng T Boylen CT Nielsen D Richman DD Tilles JG Leedom J McCutchen JA A controlled trial of early adjunctive treatment with corticosteroids for Pneumocystis carinii pneumonia in the acquired immunodeficiency syndrome. California Collaborative Treatment Group N Engl J Med 1990 323 1451 1457 2233917
Jeantils V Nguyen G Bacle F Thomas M Adjunctive treatment with corticosteroids for Pneumocystis carinii pneumonia in AIDS Therapie 1993 48 70 71 8356560
Montaner JS Guillemi S Quieffin J Lawson L Le T O'Shaughnessy M Ruedy J Schechter MT Oral corticosteroids in patients with mild Pneumocystis carinii pneumonia and the acquired immune deficiency syndrome (AIDS) Tuber Lung Dis 1993 74 173 179 8369511 10.1016/0962-8479(93)90007-K
Walmsley S Levinton C Brunton J Muradali D Rappaport D Bast M Spence D Salit I A multicenter randomized double-blind placebo-controlled trial of adjunctive corticosteroids in the treatment of Pneumocystis carinii pneumonia complicating the acquired immune deficiency syndrome J Acquir Immune Defic Syndr Hum Retrovirol 1995 8 348 357 7882099
Fisk DT Meshnick S Kazanjian PH Pneumocystis carinii pneumonia in patients in the developing world who have acquired immunodeficiency syndrome Clin Infect Dis 2003 36 70 78 12491205 10.1086/344951
Sepkowitz KA Opportunistic infections in patients with and patients without Acquired Immunodeficiency Syndrome Clin Infect Dis 2002 34 1098 1107 11914999 10.1086/339548
Lambertus MW Goetz MB Murthy AR Mathisen GE Complications of corticosteroid therapy in patients with the acquired immunodeficiency syndrome and Pneumocystis carinii pneumonia Chest 1990 98 38 43 2361410
Nelson MR Erskine D Hawkins DA Gazzard BG Treatment with corticosteroids--a risk factor for the development of clinical cytomegalovirus disease in AIDS AIDS 1993 7 375 378 8097096
Gallant JE Chaisson RE Moore RD The effect of adjunctive corticosteroids for the treatment of Pneumocystis carinii pneumonia on mortality and subsequent complications Chest 1998 114 1258 1263 9823998
Pareja JG Garland R Koziel H Use of adjunctive corticosteroids in severe adult non-HIV Pneumocystis carinii pneumonia Chest 1998 113 1215 1224 9596297
|
16271157
|
PMC1309617
|
CC BY
|
2021-01-04 16:28:17
|
no
|
BMC Infect Dis. 2005 Nov 7; 5:101
|
utf-8
|
BMC Infect Dis
| 2,005 |
10.1186/1471-2334-5-101
|
oa_comm
|
==== Front
BMC PsychiatryBMC Psychiatry1471-244XBioMed Central London 1471-244X-5-421628392610.1186/1471-244X-5-42Research ArticleBright green light treatment of depression for older adults [ISRCTN69400161] Loving Richard T [email protected] Daniel F [email protected] Nancy C [email protected] Michael A [email protected] Department of Psychiatry, University of California, San Diego, USA2005 9 11 2005 5 42 42 13 5 2005 9 11 2005 Copyright © 2005 Loving et al; licensee BioMed Central Ltd.2005Loving et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
Bright white light has been successfully used for the treatment of depression. There is interest in identifying which spectral colors of light are the most efficient in the treatment of depression. It is theorized that green light could decrease the intensity duration of exposure needed. Late Wake Treatment (LWT), sleep deprivation for the last half of one night, is associated with rapid mood improvement which has been sustained by light treatment. Because spectral responsiveness may differ by age, we examined whether green light would provide efficient antidepressant treatment in an elder age group.
Methods
We contrasted one hour of bright green light (1,200 Lux) and one hour of dim red light placebo (<10 Lux) in a randomized treatment trial with depressed elders. Participants were observed in their homes with mood scales, wrist actigraphy and light monitoring. On the day prior to beginning treatment, the participants self-administered LWT.
Results
The protocol was completed by 33 subjects who were 59 to 80 years old. Mood improved on average 23% for all subjects, but there were no significant statistical differences between treatment and placebo groups. There were negligible adverse reactions to the bright green light, which was well tolerated.
Conclusion
Bright green light was not shown to have an antidepressant effect in the age group of this study, but a larger trial with brighter green light might be of value.
==== Body
Background
Bright white light has been shown to suppress melatonin, shift circadian rhythms and alleviate depression. Evidence suggests that green light may have effects similar to those of white light but could be more efficient [1-4]
Recent studies have suggested that lower photon densities of blue light are required to suppress melatonin or to shift circadian phase than green, yellow, or red light [5-8]. It has been suggested that ganglion cells supplying the retinohypothalamic tract may be particularly responsive to blue light because they contain the blue-light-absorbing photopigment, melanopsin [9,10]. Retinohypothalamic transmission is key to light suppression of melatonin and to light-induced circadian phase shifting. However, there are reasons for skepticism that blue light is the best choice for light treatment of depression. The role of the retinohypothalamic tract in the antidepressant effects of bright light has not been fully established. The retinohypothalamic tract can be stimulated by light in the absence of melanopsin, suggesting an ancillary role for rods or cones which may be more sensitive to green light [11]. Indeed, the actual empirical peak wavelength for stimulation of melanopsin cells was about 500 nm (blue-green), although fitting an opsin curve to the data yielded interpolated blue peaks around 470–480 nm [12,13]. A single opsin curve cannot be accurately explanatory if melanopsin, rods, and cones are all involved. In some studies, adjustment of retinal sensitivity spectra for attenuation of blue light by the ocular lens may have suggested a greater advantage for blue light than was empirically observed with corneal illumination [5]. Indeed, in older subjects, whose yellowing lens increasingly attenuates blue light, green light might be as effective as blue light in suppressing melatonin [14-16]. The practical advantage of blue light may also be exaggerated by expressing sensitivity in photon densities, since blue photons contain more energy as compared to green or red. The risks of blue light are greater [17]. For these considerations, the benefits/risks ratio could be better with green light than with blue light, possibly better than with most white light. Previous studies of green light have suggested positive benefits with modest brightness, e.g., 200–2500 lux [4,18,19].
The mood-improving effects of one night of partial sleep deprivation are enhanced by subsequent daily use of light treatment. Our laboratory's earlier studies, [20,21] the Praskos' study, [22] and the work of Neumeister et al., [23] Loving, [24] and Bloching [25] all indicated that Late Wake Treatment (LWT), when combined with bright light, produces remarkable antidepressant responses, demonstrated by dramatic contrasts between bright light and placebo. Considering the evidence that LWT may accentuate the contrast of bright light and placebo, we believed LWT would add to the potential light treatment effect anticipated in this study.
In a previous study to be reported separately, we found no significant benefit of bright white light for elderly depressed outpatients . This study sought to determine if light resistance in the 60–79 year age range could be overcome with bright green light. Specifically, we ran a 4-week clinical trial of morning bright green light versus dim red placebo light, as an adjunctive antidepressant treatment.
Besides changing the active treatment from 8,500 lux white fluorescent light to 1,200 lux green LED light, this study differed from our previous clinical trial of depressed elders by focusing on morning treatment regardless of the subject's chronotype (circadian adjustment) and by randomizing active treatment immediately after intake, without baseline recording. In this study, no hormone data were collected.
Methods
Recruitment for depressed individuals, age 60–79 years, was conducted by advertising and public presentations, from February 2004 to January 2005. Written informed consent was obtained from each participant in accordance with the guidelines set forth by the Declaration of Helsinki. The study protocol and consent form were approved by the UCSD Human Research Protection Program. In addition, those participants who were being treated for depression by either a physician or counselor were requested to obtain the written agreement of the therapist for the study, to assure that there was no interference with ongoing treatment and treatment responsibility. Patients were encouraged to continue ongoing treatment during the study, with the assumption that psychotherapy and medication effects over an interval of 4 weeks were likely to be small and randomized between groups.
For enrollment in the study, a Geriatric Depression Scale (GDS) score of 11 (indicating probable major depression) [26] was required. Meeting full DSM-IV criteria for current major depressive disorder was not required, because many aging depressed people need treatment without meeting criteria for major depressive disorder [27]. As a history of mania appears to predict a greatly increased risk of a manic switch during bright light treatment, any volunteer with a history of mania was excluded [28]. Volunteers who were outdoors so much that artificial light treatment would add little were also excluded. For example, if they were outdoors for more than an hour at times of potential light triggered circadian rhythm shifts, that is during morning or evening hours. The Structured Clinical Interview for DSM-IV Axis I Disorders (Non-patient Edition) [29] was administered during the assessment period.
Subjects were then randomized into one of two treatment groups: A) 1,200 lux bright green light or B) 10 lux dim-red-light placebo. Computerized randomized assignment within blocks, using sealed envelopes, was stratified for age below or ≥ 68 years, and baseline GDS score below or ≥ 16. One hour of treatment was self-administered within one hour of awakening. Subjects were instructed to place the light box at eye level and to sit so their eyes were 24" from the light box. Both red (SunBox Company, Gaithersburg, MD) and green (Apollo Health, Orem, UT) light boxes were specially built for the study. They each used light emitting diodes (LED) as the light source (See Figure 1). Both light sources were enclosed in standard light treatment boxes, approximately 18" × 24" × 4", with a clear light diffuser panel on the front.
Figure 1 Spectrophotometric measures of illumination are shown comparing daylight with the green and red treatment lights. Sunlight was measured with the photometer pointed towards the horizon (and shaded from direct sun) near noon on a clear sunny day (32.85 North latitude, 2/2/05). Green light was measured at 2 feet with the photometer oriented towards the center of the box. The red light was measured with the photometer adjacent to the diffuser, because at 2 feet, the illumination was too dim to be plotted on the same scale. The irradiance scale was arbitrary (uncalibrated) but identical for the three measures.
We employed dim red light as a placebo, reasoning that because the light was dim and because the red part of the spectrum may be relatively inactive biologically [5] there would be no substantial biological effect. Fortunately, claims by others of red-light benefits allowed us to tell volunteers, without deception, that some people think that red light is active, even though we are skeptical. In this way, we attempted to keep subjects blind to treatment expectations.
During the four week protocol, the volunteers completed sleep-activity logs daily. They continuously wore an Actillume wrist monitor (Ambulatory Monitoring Inc., Ardsley, N.Y.) to record sleep-wake data and illumination data.
To test the benefits of partial sleep deprivation, on the initial night of the research protocol, we asked volunteers to awaken themselves 4 hours after going to bed and to remain awake for the second half of the night. They were asked to call our telephone answering machine every half hour to confirm that they had been awake for that time period. In a previous study, we found such home sleep deprivations work well without complication [24], but in our previous study of depressed elderly, few could successfully comply.
To assess subject expectations for the two treatments, measures were taken at the beginning and end of treatment, using 100 mm visual analog scales for both mood and sleep improvement. The initial rating was obtained after the subject was randomized and had seen the light they would be using but before the first actual treatment. A final assessment was obtained on the last day of the study.
Four weeks of treatment were carried out with weekly symptom assessments and continuous wrist recordings of activity and illumination exposure. The investigators visited subjects weekly to assure their safety and their compliance with the study, to administer and collect rating forms, and to transfer data from the Actillume recorders. A final symptom and circadian assessment was completed in the last 48 hours of the 4-week randomized treatment. Two-week and 4-week follow-up assessments were obtained.
In addition to daily log sheets used to record activity, sleep behaviors, and visual analog self ratings of mood, the subjects completed a weekly GDS [26] and a Systematic Assessment for Treatment Emergent Events (SAFTEE) [30] symptom scale. GDS ratings were also completed at the usual time of awakening after the half-night sleep deprivation. Further mood measurements were made at baseline and end-of-treatment using the SIGH_SAD_SR, a self-rating form including the Hamilton Depression Rating Scale (HDRS, 17 items used) and atypical items previously shown to be responsive to light treatment [31]. Additionally, when a graduate student was available, a blind HDRS rating interview was obtained at baseline and in the last week of the study.
Records from the Actillume monitor indicating total activity, sleep-wake, and log10[lux] were fitted to cosine curves for each subject. The mesors or fitted cosine means were examined, as well as the acrophases which indicate the time of day of the fitted peak.
Results
Sixty-one potential participants signed consents and were initially screened. Twenty-eight candidates did not meet GDS criteria, and 33 subjects completed the protocol, 5 males and 28 females. Bright green light treatment was received by 17 subjects, and 16 received dim red light placebo. The mean age for those completing the study was 67.7 years (SD = 6.35) and ranged from 59 to 80.
Based on the SCID interviews, DSM IV Axis I diagnoses for the sample were Major Depressive Disorder (N = 30) and Minor Depressive Disorder (N = 1). Two subjects had GDS ≥ 11 but did not meet SCID criteria for any Axis I diagnosis. None of the subjects met criteria for DSM IV Seasonal Trend. The characteristics of these depressed subjects ranged from single episodes to chronic recurrent conditions and varied in severity. This added to the general applicability of the treatments.
Of the 33 subjects who completed the study, 13 were being seen by a psychotherapist during research treatment. Of these 13, 7 received bright green light and 6 received dim light. In all 13 cases, treatment was stable during the study. Antidepressants medication was used by 13 of the subjects (5 bright green light, 8 dim light). Other prescription medications used by participants were; antianxiety (2), cardiac (7), antihypertensive (11), analgesic (6), hypnotic (8), thyroid (9), hormone replacement therapy (9), diabetes drugs (1), cholesterol-lowering compounds (12). Nine subjects received both psychotherapy and psychiatric medication, of whom 5 received bright green light and 4 received dim light. The 5 subjects receiving bright green light had no medication changes, and of the 4 dim light subjects, 2 had minor medication changes, one increased fluoxetine from 20 mg to 40 mg per day at week 2 and one reduced citalopram from 20 mg to 10 mg only for the second week of the study.
For the 33 subjects who completed the protocol, groups assigned to active and control light treatment were balanced in age and severity of depression (see Table 1). Expectations for sleep and mood effects of light treatment are shown in Table 2. A repeated-measures MANOVA as well as a one-way ANOVA of change scores showed no difference in the sleep and mood expectations between the bright and dim treatment groups either before or after treatment, suggesting that the dim red light was an active control which balanced expectations.
Table 1 Stratification by Age and Depression Severity
Age-Depression Severity Group Bright Dim Total
Age < 68, GDS < 16 0 0 0
Age < 68, GDS ≥ 16 10 10 20
Age ≥ 68, GDS < 16 2 2 4
Age ≥ 68, GDS ≥ 16 5 4 9
Total 17 16 33
Table 2 Expectations for Improvement in Sleep and Mood
100 mm Visual Analog Scale, 0 = Worse 100 = Better
Measure Mood Sleep
Light Green Red Green Red
Time Initial Final Initial Final Initial Final Initial Final
N 16 16 14 14 16 16 14 14
Mean 76.3 69.6 83.4 70.1 71.7 71.8 78.9 64.1
SD 18.0 13.6 12.5 20.0 20.9 9.7 12.9 24.0
Sleep changes
Sleep parameters were not statistically different between the green light treatment group and the red light placebo group. There was no significant effect on sleep onset time or sleep offset time, nor did total sleep time vary significantly by treatment. The total amount of Actillume-estimated sleep was balanced at baseline and not significantly affected by treatment assignment. The baseline, initial week, estimate of total sleep during the nocturnal period was 332 minutes. During the final week the estimate of total sleep during the nocturnal period was 347 minutes. The sleep efficiencies for the initial week and the final week were 75.2 and 74.5 respectively. Wake after sleep onset did not vary at baseline or at the end of treatment. The number of awakenings during the night and the sleep latency did not vary significantly either at the beginning or end of treatment.
Mood improvement
Mean mood scores for the different groups at each measurement point are shown in Table 3. Subjects' mood improved under both treatments. The average GDS score improved by 7 points (an average of 23%). There were no significant treatment differences in GDS improvement by ANCOVA. The average HDRS17 (extracted from the self-rated SIGH-SAD-SR) improved by 7 points. Improvements following wake treatment were not statistically significant. There were no significant treatment differences in HDRS17 improvement or atypical scores. Blind HDRS17 ratings, when available, were consistent with the self-rating (HDRS17) scores. Power analysis estimated 80% power to detect a large effect size of 0.51 in either the GDS or self-rated HDRS17.
Table 3 GDS and HDRS Scores by Week by Light Condition Mean (SD) N
Light Condition Bright Light Dim Light
Baseline 20.5 (5.19) 17 20.1 (3.98) 16
After Wake Treatment 16.1 (7.33) 17 18.5 (4.81) 16
Treatment week 1 19.4 (5.51) 16 19.5 (5.26) 14
Treatment week 2 15.5 (7.57) 15 15.6 (5.50) 16
Treatment week 3 13.3 (6.72) 16 15.9 (5.92) 16
Treatment week 4 12.1 (6.25) 17 14.0 (6.30) 16
Two-week follow-up 13.5 (7.53) 17 10.7 (6.89) 11
Four-week follow-up 12.4 (7.66) 16 10.3 (7.07) 10
3-Month follow-up 11.6 (7.59) 13 15.8 (6.94) 6
HDRS 17 Self Report – Baseline 19.1 (4.13) 17 17.0 (4.80) 16
HDRS 17 Self Report – Final 11.4 (3.94) 17 10.8 (5.01) 16
Blind HDRS 17 – Baseline 18.0 (4.87) 13 17.9 (6.34) 14
Blind HDRS 17 – Final 11.4 (5.11) 13 10.5 (6.73) 14
Adverse reactions
Participants experienced no psychiatric hospitalizations, suicide attempts, or deaths during the study or follow-up. There were no incidents of mania or hypomania during the light treatment or during follow-up.
The weekly SAFTEE physical symptom inventory was examined for adverse reactions to both of the light treatments. Ninety-six individual symptoms were evaluated for change during the light treatment period. To improve the stability of measurement, symptoms were grouped into 17 SAFTEE-defined categories, which were tested with Wilcoxon's Signed Rank Test. The results of these group tests are contained in Table 4. The symptom groups for Eyes, Urination, Muscle/Bone, and Other all improved in the bright green light condition. The symptom groups for Mouth/Teeth, Gynecology, and Muscle Bone improved in the dim red light condition. No symptom groups worsened under either light condition The "Other" category contains a large number of mood related items. There were no significant contrasts in SAFTEE changes between treatments.
Table 4 SAFTEE Symptoms, Mean Scores for Beginning and End of Light Treatment with Wilcoxon Signed Ranks Test
Light Condition Bright Dim
Symptom Category Baseline Post Treatment p Baseline Post Treatment p
Head 5.62 4.94 .078 5.31 5.25 .760
Eyes 9.41 7.41 .009 10.19 8.38 .212
Ears 5.12 5.06 .739 5.94 6.06 .661
Mouth/Teeth 8.18 7.53 .754 9.19 8.31 .043
Nose/Throat 7.35 6.59 .052 7.62 6.81 .283
Chest 8.64 8.18 .231 7.81 7.88 .666
Heart 2.18 2.53 .063 2.12 2.62 .480
Stomach/Abdomen 5.88 5.59 .474 5.38 5.69 .144
Bowel 8.35 7.59 .192 9.62 8.62 .305
Appetite 6.76 6.47 .475 7.50 6.62 .117
Urination 7.71 6.76 .042 7.12 7.25 .856
Gynecology 3.47 3.71 1.00 5.19 2.25 .026
Genital/Sexual 5.06 5.35 .799 6.38 4.69 .074
Muscle/Bone 7.47 6.00 .011 6.25 4.81 .016
Walking/Moving 9.24 8.12 .065 8.62 8.62 .152
Scalp/Skin 5.82 5.76 .887 6.44 5.81 .232
Other 33.76 24.12 .001 27.81 23.06 .052
Acrophase changes
The two treatment groups had no significant differences in Actillume-recorded activity, sleep, or light exposure at baseline (see Table 5), that is, they were essentially identical in these measures. The fitted peaks (acrophases) of the 24-hour rhythms of activity, sleep, and light exposure were all later in the last week of treatment than the first week, but no contrasts between the bright green and dim red light treatments were significant (see Table 6).
Table 5 Baseline Activity, Light, and Sleep Measures by Light Condition
ACTIVITY LIGHT SLEEP
Baseline MESOR ACROPHASE HH:MM MESOR ACROPHASE HH:MM MESOR ACROPHASE HH:MM
Green Mean 12.42 14:05 1.08 13:29 .27 02:49
N 16 16 16 16 16 16
S.D. 3.77 02:13 .22 01:26 .06 01:46
Dim Red Mean 12.52 14:39 1.03 14:07 .27 03:24
N 16 16 16 16 16 16
S.D. 3.80 01:38 .25 01:33 .05 01:49
p (ANOVA) .939 .425 .488 .233 .775 .373
Table 6 Acrophase Changes (minutes) by Light Condition Mean (SD) N
Acrophase Change (minutes) Baseline minus Final Bright green light Dim red light p (ANOVA)
Light -34.05 (52.32) 15 -5.02 (53.25) 16 0.14
Activity -49.20 (106.75) 16 -5.35 (54.30) 16 0.15
Sleep -34.63 (59.04) 16 -4.48 (67.75) 15 0.20
Note: Negative change indicates a later (delayed) acrophase at the end of treatment.
Although an objective measure of compliance was not available, examination of the light records from the Actillume data suggested a high degree of compliance with treatment. The self reports of treatment time and duration were almost entirely consistent with instructions. The absence of dropouts and the compliance of participants with wearing Actillumes and completing logs and questionnaires also suggested that compliance with treatment was probably high.
Green light acceptance
The bright green light boxes were well tolerated by the subjects. There was no need to reduce intensity as there sometimes is with bright white light. There was no complaint of eye strain or discomfort. Some of the subjects who had experience with bright white light either before or after the study expressed a preference for the green light.
Discussion
Bright green light treatment did not increase symptom complaints in the 60–79 year old age group and caused only negligible adverse effects. Although no statistically significant advantages of the 1,200 lux green light were found, the trend in self-ratings favored the green light, and the effect sizes on self-rating depression scales were comparable to those in many trials of 8+ weeks of antidepressant drugs [32]. Most antidepressants drug trials are planned on a larger scale to achieve significance even with small effect sizes (which may be all that the antidepressant achieves). The current study only had power to detect a large effect, which was not observed.
Our observations of the benefits of wake treatment were comparable to the findings in our previous study, again suggesting that wake therapy might not be effective in this age group. Part of the reason is certainly the difficulty of people in this age group in successfully remaining awake for the required interval. It has also been suggested that sleep deprivation may actually interfere with antidepressant treatment for the elderly, [33] which may possibly explain the lack of success of this trial.
Treatments were balanced by age and severity of depression, validating that the stratified randomization procedures were successful. The balancing of participant expectations and the similarity of blind and self-rated HDRS-17 scores indicate that the difficulties in blinding light treatment probably had no influence on the outcome.
There are several possible explanations for the antidepressant effects found from both of the light treatments in this study. The placebo effect must be considered in all clinical research. There were positive expectations for both treatments for both sleep and mood. The socialization and daily structure provided by the study may have lead to improved mood. Other trials have supported the positive effect of green light, but not in this age group. Yellowing of the lens in some elders may tend to attenuate light of 500 nm, though to a lesser extent than shorter wavelengths [15]. The lack of a more distinct advantage for the green light in this study may have been due to inadequate intensity and/or duration of treatment or partial light resistance in this age group. The failure to demonstrate phase advances in actigraphic measures with the green light would be consistent with inadequate green light intensity, light resistance, or poorer compliance than we have otherwise appreciated.
In this our first study with LED green light treatment, we selected a modest intensity of treatment, considerably less than the green component of sunlight and much less than broad-band visible illumination of sunlight (Figure 1). On the one hand, successful white light treatment has often required much more than 1,200 lux. On the other hand, even 400 lux of green light augmented citalopram treatment in a previous study, [19] though a limitation of that study was a possible confounding of the light effect with an earlier waking time. Since the green light was so well-tolerated, a larger-scale trial with somewhat brighter green light would appear safe and might well demonstrate a significant benefit.
Conclusion
This trial of moderate bright green light contrasted with dim red light placebo did not demonstrate significant difference in mood, sleep or activity measures in this age group. The trail of Late Wake Treatment similarly, did not produce significant improvement in mood measures.
Competing interests
The author(s) declare that they have no competing interests.
Authors' contributions
RTL coordinated and carried out the clinical trial, performed statistical analyses, and drafted the manuscript. DFK conceived and drafted the design, administered and participated in data collection, and participated in statistical analyses and manuscript preparation. NCK carried out subject recruitment and collection and scoring of Actillume recordings and subject questionnaires. MAG constructed a software database and administered blind HDRS ratings. All authors reviewed and approved the final manuscript.
Pre-publication history
The pre-publication history for this paper can be accessed here:
Acknowledgements
The authors acknowledge Sunbox of Gaithersburg, Maryland and Apollo Health of Orem, Utah for donation of the light fixtures used in this research. Jeffrey A. Elliott contributed design considerations and reviewed the manuscript. Appreciation and acknowledgment is extended to the volunteers without whom this research would not have been possible. Supported by the NIH National Institute On Aging (Ag12364, MH68545, and HL071123
==== Refs
Horne JA Donlon J Arendt J Green light attenuates melatonin output and sleepiness during sleep deprivation Sleep 1991 14 233 240 1896724
Riesenberg T Kripke DF Elliott JA Cole R Green light suppression of melatonin [abstract] Chronobiol Int 2003 20 1183
Geetha L Subbaraj R Green light evokes maximum phase shifts in the locomotor activity rhythm of the field mouse Mus booduga Journal of Photochemistry & Photobiology B – Biology 1996 33 79 82 10.1016/1011-1344(95)07224-1
Oren DA Brainard GC Johnston SH Joseph-Vanderpool JR Sorek E Rosenthal NE Treatment of seasonal affective disorder with green light and red light Am J Psychiatry 1991 148 509 511 2006698
Brainard GC Hanifin JP Greeson JM Byrne B Glickman G Gerner E Rollag MD Action spectrum for melatonin regulation in humans: Evidence for a novel circadian photoreceptor J Neurosci 2001 21 6405 6412 11487664
Thapan K Arendt J Skene DJ An action spectrum for melatonin suppression: evidence for a novel non-rod, non-cone photoreceptor system in humans J Physiol (London) 2001 535 261 267 11507175 10.1111/j.1469-7793.2001.t01-1-00261.x
Wright HR Lack LC Effect of light wavelength on suppression and phase delay of the melatonin rhythm Chronobiol Intl 2001 18 801 808 10.1081/CBI-100107515
Lockley SW Brainard GC Czeisler CA High sensitivity of the human circadian melatonin rhythm to resetting by short wavelength light The Journal of Clinical Endocrinology & Metabolism 2003 88 4502 4505 12970330 10.1210/jc.2003-030570
Hattar S Lucas RJ Mrosovsky N Thompson S Douglas RH Hankins MW Lem J Biel M Hofmann F Foster RG Melanopsin and rod-cone photoreceptive systems account for all major accessory visual functions in mice Nature 2003 424 76 81 12808468 10.1038/nature01761
Panda S Provencio I Tu DC Pires SS Rollag MD Castrucci AM Pletcher MT Sato TK Wiltshire T Andahazy M Melanopsin is required for non-image-forming photic responses in blind mice Science 2003 301 525 527 12829787 10.1126/science.1086179
Panda S Sato TK Castrucci AM Rollag MD Degrip WJ Hogenesch JB Provencio I Kay SA Melanopsin (Opn4) requirement for normal light-induced circadian phase shifting Science 2002 298 2213 2216 12481141 10.1126/science.1076848
Berson DM Dunn FA Takao M Phototransduction by retinal ganglion cells that set the circadian clock Science 2002 295 1070 1074 11834835 10.1126/science.1067262
Dacey DM Liao H-W Peterson BB Robinson FR Smith VC Pokorny J Yau K-W Gamlin PD Melanopsin-expressing ganglion cells in primate retina signal colour and irradiance and project to the LGN Nature 2005 433 749 754 15716953 10.1038/nature03387
Herljevic M Middleton B Thapan K Skene DJ Light-induced melatonin suppression: age-related reduction in the response to short wavelenth light [abstract] Society for Research on Biological Rhythms, Program and Abstracts 2004 124
Charman WN Age, lens transmittance, and the possible effects of light on melatonin suppression Ophthal Physiol Opt 2003 23 181 187 10.1046/j.1475-1313.2003.00105.x
Wright HR Lack LC Kennaway DJ Differential effects of light wavelength in phase advancing the melatonin rhythm J Pineal Res 2004 36 140 144 14962066 10.1046/j.1600-079X.2003.00108.x
Sliney DH Holick MF, Kligman AM The potential ocular hazards of viewing bright light sources Biologic Effects of Light 1992 Berlin: Walter de Gruyter 230 244
Ahmed S Cutter NL Lewy AJ Bauer VK Sack RL Cardoza MS Phase response curve of low-intensity green light in winter depressives [abstract] Sleep Res 1995 24 508
Benedetti F Colombo C Pontiggia A Bernasconi A Florita M Smeraldi E Morning light treatment hastens the antidepressant effect of Citalopram: A placebo-controlled trial J Clin Psychiatry 2003 64 648 653 12823078
Kripke DF Risch SC Janowsky DS Bright white light alleviates depression Psychiatry Res 1983 10 105 112 6581487 10.1016/0165-1781(83)90109-9
Kripke DF Risch SC Janowsky DS Lighting up depression Psychopharmacol Bull 1983 19 526 530
Prasko J Foldmann P Praskova H Zindr V Hastened onset of the effect of antidepressive drugs when using three types of timing of intensive white light Cs Psychiat 1988 84 373 383
Neumeister A Goessler R Lucht M Kapitany T Bamas C Kasper S Bright light therapy stabilizes the antidepressant effect of partial sleep deprivation Biol Psychiatry 1996 39 16 21 8719121 10.1016/0006-3223(95)00086-0
Loving RT Kripke DF Shuchter SR Bright light augments antidepressant effects of medication and wake therapy Depression and Anxiety 2002 16 1 3 12203667 10.1002/da.10036
Bloching B Dechene C Taschner KL Outlasting antidepressant effect of late partial sleep deprivation by bright light therapy J Sleep Res 2000 9 21
Yesavage JA Brink TL Rose TL Lum O Huang V Adey M Leirer VO Development and validation of a geriatric depression screening scale: a preliminary report J Psychiatr Res 1983 17 37 49 7183759 10.1016/0022-3956(82)90033-4
Lyness JM King DA Cox C Yoediono Z Caine ED The importance of subsyndromal depression in older primary care patients: prevalence and associated functional disability J Am Geriatr Soc 1999 47 647 652 10366161
Kripke DF Timing of phototherapy and occurrence of mania Biol Psychiatry 1991 29 1156 1157 1873376 10.1016/0006-3223(91)90258-N
First MB Spitzer RL Gibbon M Williams JBW Structured Clinical Interview for DSM-IV Axis I Disorders – Non-patient Edition (SCID-I/NP, Version 20) 1996 New York City: New York State Psychiatric Institute
Moynihan C Instruction Manual for Systematic Assessment for Treatment Emergent Events (SAFTEE) 1983 Rockville, MD: ADAMHA
Williams JBW Link MJ Rosenthal NE Amira L Terman M Structured Interview Guide for the Hamilton Depression Rating Scale: Seasonal Affective Disorder Version 2002 New York, New York State Psychiatric Institute 1 9
Moncrieff J Wessely S Hardy R Active placebos versus antidepressants for depression (Cochrane Review) The Cochrane Library 2004 3 Chichester, UK: John Wiley & Sons, Ltd
Reynolds CF Smith GS Dew MA Mulsant BH Miller MD Schlernitzauer M Stack JA Houck PR Pollock BG Accelerating symptoms-reduction in late-life depression A double-blind, randomized, placebo-controlled trial of sleep deprivation Am J Geriatr Psychiatry 2005 13 353 358 15879583 10.1176/appi.ajgp.13.5.353
Loving RT Kripke DF Elliot JA Knickerbocker NC Grandner MA Bright light treatment of depression for older adults BMC Psychiatry
|
16283926
|
PMC1309618
|
CC BY
|
2021-01-04 16:33:02
|
no
|
BMC Psychiatry. 2005 Nov 9; 5:42
|
utf-8
|
BMC Psychiatry
| 2,005 |
10.1186/1471-244X-5-42
|
oa_comm
|
==== Front
Nutr Metab (Lond)Nutrition & Metabolism1743-7075BioMed Central London 1743-7075-2-291623231510.1186/1743-7075-2-29ReviewPrevalence, predisposition and prevention of type II diabetes Cheng Dong [email protected] Department of Obesity and Metabolic Research, Pharmaceutical Research Institute, Bristol-Myers Squibb Company, Princeton, NJ 08543, USA2005 18 10 2005 2 29 29 24 8 2005 18 10 2005 Copyright © 2005 Cheng; licensee BioMed Central Ltd.2005Cheng; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
In 2000, more than 151 million people in the world are diabetic. It is predicted that by 2010, 221 million people and by 2025, 324 million will be diabetic. In the U.S., for the population born in 2000, the estimated lifetime risk for diabetes is more than 1 in 3. The economic and human cost of this disease is devastating. The current cost of diabetes in the U.S. is estimated to be at $132 billion, which includes $92 billion of direct medical costs and $40 billion of indirect costs such as disability, work loss and premature mortality. The outbreak of the current diabetic epidemic has been accompanied by a similarly drastic increase in obesity. The relation between the two is a matter of debate but presumably both are caused by changes in dietary habits and an increasingly sedentary modern lifestyle. Compelling scientific evidence indicates that lifestyle modification effectively prevents or delays the occurrence of type 2 diabetes. Recent clinical trials also demonstrate that success in the treatment of obesity, either surgically or pharmacologically, leads to the prevention of type 2 diabetes among the obese. Clinical data have also revealed that the insulin sensitizing agent troglitazone is efficacious in both β-cell preservation and delaying the onset of type 2 diabetes. Future safe and more effective anti-obesity medicines and insulin sensitizing agents that help to preserve β-cell function, in addition to efforts of lifestyle modification, thus hold promise for the overweight population with potential for reduction in the development of diabetics.
==== Body
Background
Social affluence is a double edged sword. On the one hand, life is more convenient than ever because of the advances in technology; on the other hand, the incidence of diabetes is occurring at an alarming rate. The explosive increase in number of people diagnosed with diabetes makes this disease a new health threat in the 21st century. Understanding the etiology of and finding a way to prevent diabetes, especially type 2 diabetes, is an urgent challenge for the health care community and our society.
Epidemics of type 2 diabetes
In 2002, more than 18 million Americans, about 6.9% of the U.S. population, have diabetes [1]. Globally, the number of people that has been diagnosed with diabetes has also exploded in the past two decades. In 2000, 151 million people in the world were diabetic. With the current rate of increase, it has been projected that 221 million people will be diabetic in 2010 and 324 million by 2025 [2].
There are two major forms of diabetes: type 1 and type 2 diabetes [3]. The hall mark of type 1 diabetes is the destruction of insulin producing β-cells in the pancreas, primarily due to autoimmune responses. Type 1 diabetes is manifested with absolute insulin deficiency. In contrast, type 2 diabetes is characterized by two defects: insulin deficiency and insulin resistance. Type 2 diabetes accounts for 90 to 95% of the incidence of diabetes. The current epidemic outbreak of diabetes reflects the high prevalence of type 2 diabetes.
While the seriousness of the epidemic was only fully recognized recently, the threat by a trend of the increase of incidences in diabetes was first recognized by Elliott Joslin some 80 years ago [4]. From a historical perspective, the epidemic of type 2 diabetes today has developed steadily through the decades in the last century. According to the data of National Health Interview Survey, the incidence in 1990–1992 was 6.4 times the rate of 1935–1936 [5]. In a 10-year span from 1990 to 1999, the prevalence further increased by 40% from 4.9% to 6.9% [6]. Assuming the rate of increase in incidences continues, the data from the National Health Interview Survey (1984–2000) predicts that the residual lifetime risk of diabetes for individuals born in 2000 is 32.8% for males and 38.5% for females. The highest risk for ethnic subpopulations is in Hispanic females whose lifetime risk of becoming diabetic is 52.5% [6].
The global statistics indicate that the burden of type 2 diabetes is not restricted to the developed nations, but also is a problem for developing countries. For example, in the Pacific island of Nauru, type 2 diabetes is present in about 40% of adults [7]. Ironically this disease was not present a half century ago in this region. In a 11-year follow-up study in Mauritius, the prevalence of type 2 diabetes, in both men and women, regardless of ethnic group, increased steadily from 12.8% in 1987, to 15.2% in 1992, and to 17.9% in 1998 [8]. In recent years, the prevalence of type 2 diabetes has also increased in world's two largest populated nations, China [9] and India [10]. India now has the most people – 38 million – with diabetes. China is ranked second and has 23 million diabetics. By 2025, these numbers are expected to be doubled to about 73 million and 46 million, respectively [11].
Although previously type 2 diabetes was predominantly diagnosed in middle-aged or older people, the age of onset of this disease has decreased. Japan has seen an approximately 4-fold rise in the incidence of type 2 diabetes in 6- to 15-year-olds. This rate now is outnumbering that of type 1 diabetes in that country [11]. Data from the U.S. indicates that 8–45% of recently diagnosed cases of diabetes in the young is due to type 2 diabetes [12].
Uncontrolled diabetes leads to other serious medical complications including a substantial increase in premature morbidity and mortality [13,14]. It is reported that the incidence of cardiovascular disease among adult diabetics is 37.2%, which is much higher than the incidence in the general population [14]. Prevalence of ischemic heart disease among the diabetics 18 to 44 years of age is 14 times for those without diabetes [14]. The high blood glucose of diabetes also causes both micro and macro vascular damages. Damages in large vessels lead to stroke and cardiovascular complications, whereas damages to vessels in the extremities, eyes and kidneys lead to amputation, blindness and kidney failure. As a result, each year, as many as 24,000 diabetics in the U.S. become blind [14]; more than 100,000 require kidney dialysis or kidney transplantation which accounts for more than 40% of new cases of end-stage renal disease. Furthermore, 82,000 diabetics need to have a toe or a leg amputated, accounting for more than half of all non-traumatic lower-extremity amputations in the U.S. [14]. The diabetes epidemic is thus a large assault on the health care system and a huge burden for the society. The current cost of diabetes in the U.S. is estimated as $132 billion, which includes direct medical costs of $92 billion and indirect costs (disability, work loss, premature mortality) of $40 billion [1].
Obesity, Metabolic Syndrome, prediabetes
Obesity
What is the driving force for the current worldwide epidemic of diabetes? Environmental factors such as adoption of a sedentary lifestyle, changes in eating habits and consequent obesity, are likely the main causes or at least a parallel problem. This hypothesis is supported strongly by the studies in migrating populations. For example, the prevalence of diabetes in the urban regions of India is increasing dramatically in affluent migrant Indians [15]. Rates among Asian Indians in countries such as South Africa, the U.K. and Fiji are much higher than reported in most parts of India itself [16]. The incidence is also rising among Africans who have either urbanized or immigrated to the U.S. [17,18]. Before the 1990s, the rates of type 2 diabetes in Japan were quite low; but high rates were found in Japanese living in the U.S. [16]. Rates of increased incidence of diabetes in populations of Chinese origin also vary with environment changes. In the Northern part of mainland China, in the age group 30 to 64, about 1% was the rate for type 2 diabetes; this number went to 4 to 5% among the Chinese in Singapore and to 11 to 12% in Mauritus [19]. Many native American Indian tribes have a higher prevalence of diabetes as compared with the general U.S. population; but Eskimos appears to be an exception, whose diabetes rate is not distinct from white Americans [20]. All these population-based studies reveal a similar pattern: environmental changes and adaptation of sedentary lifestyle, resulting from industrialization and migration to urban cities, lead to the development of type 2 diabetes.
Parallel to the increase of incidences in type 2 diabetes is the increase of obesity in the population. According to the definition recommended by the World Health Organization (WHO) expert committee for the classification of overweight and obesity, today, close to 65% of the U.S. adult population is overweight, and among them, above 30% are obese [21]. In this classification, Body Mass Index (BMI) between 25 to 30 is considered as overweight and BMI above 30 is considered as obese. Based on the statistics collected by National Health and Nutrition Examination Survey (NHANES), in a representative sample of the U.S. population, the prevalence of an overweight condition has increased by 40% (from 46% in the period of 1976–1980 to 64.5% in the period of 1999–2000), and the prevalence of obesity has increased by 110% (from 14.5% to 30.5%) [21,22]. In addition, the increase in weight in the young population also exhibits an alarming growth. If overweight is defined as at or above the 95th percentile of the sex-specific BMI for age growth charts, among those aged 2 through 19 years, the prevalence of overweight was 15.5% among 12-through 19-year-olds, 15.3% among 6-through 11-year-olds, and 10.4% among 2-through 5-year-olds according to 1999–2000 NHANES data. These numbers are significantly increased from 10.5%, 11.3%, and 7.2%, respectively, as collected in 1988–1994 (NHANES III).
Despite the on-going debate as to whether obesity should be labeled as a disease, obesity, directly and indirectly, has become the global health challenge. It has been estimated that close to 300,000 deaths each year in the U.S. may be attributable to obesity [23]. In a study comparing health care spending on obese and normal-weight Americans between 1997 and 2001, it was found that per capita spending for the obese was more than $1000, or 37%, higher than spending on normal-weight people in 2001 [24].
Obesity has become a burden for society, as it is a major cause of lost productivity. The financial burdens that are consequences of obesity are associated with the treatment needed for the accompanying disease states, such as cardiovascular disease, cancer, hypertension, and most intimately, to type 2 diabetes [25]. In a 16-year follow-up study of 84,941 women, it was documented that 3300 new cases of type 2 diabetes were diagnosed from 1980 to 1996 [26]. Among these cases, body weight was the single most important predictor of diabetes. As high as ~80% of the cases of type 2 diabetes could be attributed to the combined effect of inactivity and high body weight [27].
In the realm of lifestyle, the change of dietary composition, in addition to the total caloric intake, is probably another important factor that is implicated in the epidemics of type 2 diabetes. For example, one important but not well-appreciated dietary change has been the substantial increase in refined carbohydrates especially simple carbohydrate from high intake of sucrose and high fructose corn syrup, a common sweetener used in the food industry. High influx of carbohydrate is predicted to drive the increase of de novo hepatic lipogenesis and perturbs the glucose homeostasis that appear to underlie the induction of insulin resistance [28,29].
The exact molecular and cellular connection between obesity and type 2 diabetes has not been entirely explained. In particular, there is no unifying hypothesis that explains the various states of "garden-variety" insulin resistance associated with diet-induced obesity. One of the hypotheses highlights the pathological roles of lipid abnormality accompanying obesity or high body weight, postulates that accumulation of fatty acids or fatty acid derivatives in muscle and liver produce insulin resistance [30]. At the cellular level, it has been proposed that the accumulation of diacylglycerol might activate protein kinase C (PKC) (probably by PKC-θ in rodents and by PKC-β or -δ in humans) [31]. The activated PKC in turn phosphorylates and activates other serine kinases such as IKK-β[32]and JNK-1 [33], leading to phosphorylation of serine sites on IRS-1 insulin receptor substrate (IRS) 1 and 2 at Serine/Threonine sites, blunting the insulin signaling pathway[34]. It is also suggested that ceramide, a down stream product of fatty acid metabolism, down regulates insulin signaling in muscle [35]. It has been shown that ceramide blocks insulin-stimulated tyrosine phosphorylation of IRS-1 and its subsequent recruitment and activation of PI3 kinase. The abnormalities in fat metabolism and insulin resistance seem to form a vicious cycle. Hyperinsulemia from elevated caloric or carbohydrate intake especially in the presence of high fat may cause insulin resistance in adipocytes leading to elevated fatty acids which, in turn may cause insulin resistance in muscle. At the same time, insulin resistance further exacerbates the abnormalities in hepatic fat metabolism [36,37]. In addition to the adverse metabolic consequences in insulin sensing tissues, fat accumulation also has harmful effects in insulin producing β-cells. The increased tissue levels of fatty acyl CoA cause β-cell abnormalities in nondiabetic obese patients and ultimately result in obesity-dependent diabetes [38]. Of note, these potential mechanisms are not mutually exclusive. In fact, they could all play roles at various stages for the development of type 2 diabetes, a long and gradual process [39].
Metabolic Syndrome
Obesity and insulin resistance often coexist along with other abnormalities such as hypertension and dyslipidemia. In 1988, Reaven introduced the concept of Metabolic Syndrome X in the Banting Medal address, in order to emphasize the coexistence of multiple metabolic abnormalities. This term underscores the fact that insulin resistance and its compensatory hyperinsulinemia develop with dysregulation of glucose and lipid metabolism [40]. The dysregulation of glucose metabolism is represented by varying degree of glucose tolerance. The dysregulation of metabolism of lipid is manifested by the increase of plasma triglyceride levels and the decrease of plasma HDL cholesterol. Numerous population-based studies have described the characteristics of Metabolic Syndrome X. Considerable information has evolved from the original recognition of the clustering of metabolic abnormalities centering around insulin resistance/hyperinsulinemia. The original term Metabolic Syndrome X has become a synonym of Insulin Resistance Syndrome or Metabolic Syndrome. Currently, the abnormalities clustered with insulin resistance include some degree of glucose intolerance (either manifested as elevated fasting glucose or impaired glucose tolerance), dyslipidemia (elevated triglycerides, reduced HDL-c, reduced LDL-particle diameter), endothelial dysfunction (increased mononuclear cell adhesion, cellular adhesion molecules and decrease in endothelial-dependent vasodilatation), increased procoagulant factors, and elevation of inflammation [41].
The Third Report of the National Cholesterol Education Program Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) (ATP III) has provided a working definition of the syndrome for the first time. Individuals having three or more of the following criteria were defined as having the metabolic syndrome: 1) abdominal obesity: waist circumference > 102 cm in men and > 88 cm in women; 2) hypertriglyceridemia: ≥ 150 mg/dL; 3) low HDL cholesterol: < 40 mg/dL in men and < 50 mg/dL in women; 4) high blood pressure: ≥ 130/85 mm Hg; 5) high fasting glucose: ≥ 110 mg/dL [42]. World Health Organization has defined metabolic syndrome in the following way: at least 1 of abnormalities in glycemic metabolism (type 2 diabetes, impaired glucose tolerance, insulin resistance) accompanied by at least 2 other conditions: high blood pressure (≥ 140/90 mm Hg), obesity, dyslipidemia (hypertriglyceridemia or low HDL) and microalbuminuria [43]. (For comparison of the differences between the two definitions, see Table 1). The subtle differences among various definitions do not change the overall conclusion that the prevalence of metabolic syndrome is high. It has been estimated, using ATP III criteria, that the prevalence in the U.S. population was 21.8% between 1988–1994. Using 2000 census data, about 47 million U.S. residents are estimated to have the metabolic syndrome [44].
Table 1 Definition of the Metabolic Syndrome.
ATP III Definition
Any three or more of the following criteria:
a) Waist circumference: >102 cm in men, >88 cm in women
b) Serum triglycerides: ≥ 150 mg/dL
c) HDL-cholesterol: <40 mg/dL in men, < 50 mg/dL in women
d) Blood pressure: ≥ 130/85 mm Hg
e) Serum glucose: >110 mg/dL
WHO Definition
Diabetes or IFG or IGT or insulin resistance, plus at least two of the following criteria
a) Waist-to-hip ratio: >0.90 in men, >0.85 in women
b) Serum triglycerides: >150 mg/dL par or HDL-cholesterol: <35 mg/dL in men and <40 mg/dL in women
c) Blood pressure: >140/90 mmHg
d) Urinary albumin excretion rate > 20 ug/min or albumin/creatinine ratio >30 mg/g
Since the introduction of the concept of Metabolic Syndrome and the subsequent accumulation of large body of research data, it is beyond doubt that metabolic abnormalities in the areas of carbohydrate metabolism, body weight homeostasis, lipid metabolism and blood pressure regulation, tend to occur concomitantly. However, the precise criteria for the definition of Metabolic Syndrome has yet to be unified. This is probably due to the enormous complexity of physiological/pathological conditions that Metabolic Syndrome covers. As such, the clinical utility and value of diagnosing Metabolic Syndrome is debated [45,46]. In a recent Joint Statement from the American Diabetes Association and the European Association, it is stated that "Our analysis indicates that too much critically important information is missing to warrant its designation 'syndrome'. Until much needed research is completed, clinicians should evaluate and treat all CVD risk factor without regard to whether a patient meets the criteria for diagnosis of the 'metabolic syndrome"'[47].
Prediabetes
Epidemiological studies indicate that the development of type 2 diabetes takes place over a long period of time from the initial decline of insulin effectiveness ultimately progressing to frank diabetes when β-cell function collapses. In most patients, insulin resistance can be detected long before the deterioration of glucose intolerance occurs. Approximately 5 to 10% of glucose-intolerant patients progress to frank type 2 diabetes in a given year. Inasmuch as Metabolic Syndrome emphasizes the condition of insulin resistance, the syndrome itself is not type 2 diabetes, but a large percentage of the people with Metabolic Syndrome will develop type 2 diabetes if the condition of insulin sensitivity is not improved. While the confident diagnosis of Metabolic Syndrome is technically difficult because the criteria are elusive, a simpler term – prediabetes – was introduced to define patients who have modestly higher glucose levels than normal but have not yet reached diabetic glucose levels. There are two criteria for prediabetes: impaired glucose tolerance (IGT) or impaired fasting glucose (IFG). The American Diabetes Association (ADA) specifies as a 2-h postprandial glucose 140 – 199 mg/dL as IGT, and fasting plasma glucose 110 – 125 mg/dL as IFG [48]. In 2005, in the U.S. alone, 41 million adults are reported to have prediabetes [49].
The big question is when the diabetes clock starts ticking. For eye or small vessel complication, it might be the consequence of increased plasma glucose level. For cardiovascular disease, however, the risk likely starts in the prediabetes stage. Subjects with IFG have been demonstrated to have an increased risk for macrovascular disease, in a variety of population-based studies [50,51]. Thus insulin resistance has been considered as an independent risk factor for the cardiovascular disease. Because of this, the detection of IGT is of great importance for general public health, in particular to help prevent cardiovascular complications in the high risk population [52]. Although the definition of prediabetes is simpler than Metabolic Syndrome, the glucose tolerance test remains a time-consuming method. IFG can not replace the tolerance test, since each test may reflect different defective conditions of glucose metabolism. Because the glucose tolerance test is difficult to implement as a population screen test, a large number of people have been left undiagnosed and unaware of the risks they face. This is particularly true for people who are overweight. Based on NHANES III data, among the overweight adults in the U.S., 17.1% aged 45–74 years had IGT, 11.9% had IFG, 22.6% had prediabetes, and 5.6% had both IGT and IFG. It is estimated that in the year 2000, 9.1 million overweight adults aged 45–74 had IGT, 5.8 million had IFG, 11.9 million had prediabetes, and 3.0 million had IGT and IFG [53].
The genetics of type 2 diabetes and obesity
The search of human genetic factor(s) that predispose to type 2 diabetes and obesity has gone through two directions: by studying of rare mutations and by population-based gene-hunting for the primary causes [54-56]. It is generally accepted, after more than 20 years of intense effort, that environmental factors such as lifestyle and dietary composition play profound roles for the pathogenesis of type 2 diabetes and obesity. Furthermore, it is a consensus in the field that type 2 diabetes and obesity are polygenic diseases. As such, the population based gene-hunting efforts have not yielded conclusive "diabetogenes" and "obesitogenes". In contrast, through the candidate approach by studying the rare mutations, several genes have been identified as genes that are associated with type 2 diabetes and obesity. These genes and the studies are summarized in Table 2.
Table 2 Mendelian causes of type 2 diabetes and obesity
Type 2 diabetes
Mutations Gene Names Reference
ABCG8 ATP-binding cassette Subunit C [97, 98]
CAPN10 Calpain 10 [99, 100]
GCGR Glucagon receptor [101]
GCK Glucokinase [102]
KCNJ11 Potassium channel subunit J, member 11 [103]
PPARG Peroxisome proliferator-activated receptor γ [104]
HNF4A Hepatocyte nuclear factor 4α [105]
HNF1A Hepatocyte nuclear factor 1α [106]
SLC2A1 Glut 1 [107]
INS Insulin [108]
INSR Insulin receptor [109, 110]
Mitochondrial genome Mitochondrial DNA [111]
Obesity
Diabetes Leptin receptor [112]
Mc4r Melanocortin-4 receptor [113]
Obese Leptin [114]
PCSK1 Prohormone convertase 1 [115]
Pomc Pro-opiomelanocortin [116]
SLC6A14 Solute carrier family 6 member 14 [117]
Prevention of type 2 diabetes
Because of the devastating negative economic impact and human cost of type 2 diabetes, it is highly desirable that this disease can be prevented. Proof-of-concept studies for the prevention of type 2 diabetes have been conducted in several clinical trials in recent years. The outcome of these trials provide convincing evidence that either through lifestyle adjustment or through pharmacological treatment, type 2 diabetes can be prevented or, at least, be delayed. This conclusion is of profound implication in our way of thinking about the diabetes treatment and about the direction of finding new pharmacological entity as well as policy changes in the health care system.
There are four randomized controlled trials reported to-date for lifestyle modification. The conclusions from these studies are in general supportive of each other, and strongly argue for lifestyle modification as an effective way to prevent type 2 diabetes. The first study, Da Qing Study, dates back to 1986 [57]. Individuals diagnosed with IGT (577 patients) were randomized either to a control group or to one of three active treatment groups: diet only, exercise only, or diet plus exercise. Follow-up evaluation examinations were conducted at 2-year intervals over a 6-year period to identify subjects who developed type 2 diabetes. The cumulative incidence of diabetes at 6 years was 67.7%, 43.8%, 41.1%, and 46.0% in the control group, diet group, exercise group, and diet-plus-exercise group, respectively. After adjustment of BMI difference, the diet, exercise, and diet-plus-exercise interventions were associated with 31%, 46%, and 42% reductions in risk of developing diabetes, respectively.
The second study was conducted in Finland. In the Finnish trial [58], 522 middle-aged, overweight with impaired glucose tolerance subjects were randomly assigned to either the intervention group or the control group. The intervention group received individualized counseling aimed at reducing weight, total intake of fat, and intake of saturated fat and increased intake of fiber and physical activity. During the 3.2 years of follow-up, the cumulative incidence of diabetes was 11% which was reduced by 52% compared to the control group with an incidence of 23%. In this time period, the risk of diabetes was reduced by 58% in the intervention group.
In the third study, the Diabetes Prevention Program [59], a direct comparison was made between the lifestyle modification versus medication with metformin. Nondiabetic candidates (3234) with elevated fasting and post-load plasma glucose concentrations were randomized into placebo, metformin (850 mg twice daily), or a lifestyle-modification program (with the goals of at least a 7 percent weight loss and at least 150 minutes of physical activity per week). After a follow-up of 2.8 years, the incidence of diabetes was 11.0%, 7.8%, and 4.8% each year in the placebo, metformin, and lifestyle groups, respectively. The lifestyle intervention reduced the incidence by 58% and metformin by 31%, as compared with placebo; the lifestyle intervention was significantly more effective than metformin.
More recently, a fourth study comparing lifestyle modification and metformin medication was conducted in India [60]. A cohort of 531 subjects with IGT were randomized into four groups: control, lifestyle modification, metformin, lifestyle modification in addition to metformin. Diabetes developed in 49.6%, 35%, 39.8% and 34.7% respectively after 30 months of follow-up. This study adds another piece of evidence that lifestyle modification is effective in preventing diabetes. It also concluded that the lifestyle modification effect is not enhanced by metformin.
The success of these major diabetes prevention trials has impacted the way of approaching diabetes care. The ADA has recommended that all overweight people, aged ≥ 45 years with prediabetes be considered potential candidates for diabetes prevention. Overweight younger individuals with prediabetes and other risk factors should also be inducted into a diabetes prevention program [61].
Lifestyle modification
The success of the DPP and other trials provide compelling evidence that modification of lifestyle, including increasing exercise along with decreased caloric intake, is an effective way of preventing or delaying the onset of type 2 diabetes in high risk populations. Interestingly, the pharmacological agent, metformin, only provided a modest effect in the DPP trial. Furthermore, there was no additional benefit of metformin over the lifestyle modification in its effect on diabetes prevention. Since the metformin mechanism of action might be through activation of AMPK [62], a kinase that is activated during exercise [63], it is possible that exercise may act through the same molecular cellular pathway and therefore there are no additive effects of these two treatment regimens.
One important aspect of combating the epidemics of obesity and type 2 diabetes has been through dietary strategy. Recently, very-low-carbohydrate diets have gained much popularity [64-67]. These diets have produced results at least comparable to and frequently better than traditional diets [68-70] although their acceptance among official agencies is limited at best. More recently the effects of low-carbohydrate/high-protein diets on the blood glucose levels, insulin resistance have been evaluated in obese/over-weight type 2 diabetes patients. One study comprised inpatient comparisons of a low-carbohydrate diet (20 g carbohydrate per day and unlimited protein and fat) among 10 obese patients with type 2 diabetes versus another 10 patients in a usual diet for 14 days. On the low-carbohydrate diet, mean energy intake decreased from 3111 kcal/d to 2164 kcal/d. This resulted in a mean weight loss of 1.65 kg, decrease in hemoglobin A1c from 7.3% to 6.8%, and insulin sensitivity improvement by about 75% [71]. The other study used a moderately low carbohydrate diet with a carbohydrate:protein:fat ratio of 20:30:50. Ingestion of this diet for 5 weeks in patients of untreated type 2 diabetes decreased hemoglobin A1c from ~9.8% to ~7.6% [72]. These study seem to indicate, by solely decrease the carbohydrate intake, it empowers patients with diabetes to ameliorate hyperglycemia without pharmaceutical intervention. Because glucose is the major insulin secretagogue carbohydrate reduction would be expected to be beneficial in type 2 diabetes and the use of such diets has been summarized by Arora & McFarlane [64] although, as noted above, official recommendations generally continue to recommend low fat and high carbohydrate intake.
Medications in diabetes prevention
Although lifestyle modification is the most effective way of preventing diabetes in the clinical trials, the implementation demands a high level of discipline for the patients, which may preclude its effectiveness in the general population. In the long run, safe, effective medications, therefore, will likely be the best choice for intervention. One important lesson from the success of prevention trials is that lifestyle modification that results into the weight change has proven to disrupt the apparent pathological connection of overweight and the development of type 2 diabetes. Any drug that reduces weight increases the disposal of excess energy, or mimics exercise could be an effective treatment for diabetes prevention. The potential non-pharmaceutical approach and the pharmaceutical approach of diabetes prevention are summarized in table 3.
Anti-obesity drugs
The concept of using an anti-obesity strategy to prevent diabetes was dramatically demonstrated in the follow-up studies for clinically severe obese (> 45 kg excess body weight) patients that received bariatric surgery [73]. During the ~5 year follow-up, among the experimental group of patients that included 109 patients with IGT who underwent bariatric surgery for weight loss, only 1 patient developed diabetes, resulting in a conversion rate of only 0.15 cases per 100 person-years. This rate is 30 fold lower than that found in the control group where the rate was found to be 4.72 cases per 100 person-years. In another prospective study conducted in Sweden, the Swedish Obese Subjects (SOS) study, compared the incidence of development of diabetes for 346 patients receiving gastric bypass surgery to the same number of obese control subjects receiving "conventional nonpharmacological obesity treatment" [74]. For these two sets of clinically obese patients, during the 8-year follow-up, the conventional nonpharmacological obesity treatment, in the hands of a primary health care system, had no effect on body weight, whereas gastric bypass surgery resulted in 18- to 30-kg maintained weight loss. This intentional weight loss in severely obese individuals reduces the 8-year incidence of diabetes by 5-fold. In addition to the benefit illustrated in the dramatic weight loss, modest weight control also added the benefit for diabetes prevention [75,76].
Orlistat (Xenical) is one of the few existing anti-obesity drugs in the market. In 2004, after the completion of a double-blind, placebo-controlled prospective study known as Xenical in the Prevention of Diabetes in Obesity Subjects (XENDOS), the European Commission approved labelling for the reduction of risks associated with the development of type 2 diabetes. In this study, 3,305 obese patients were followed over a 4-year period. Patients who received orlistat not only lost more weight, compared with those receiving placebo (5.7 kg vs. 3.0 kg), but also exhibited a 37% reduction in the incidence of type 2 diabetes during the treatment period [77]. Additional data indicated that orlistat reduced the doses of antidiabetic drugs by 23% during the treatment [78]. However, the profound gastrointestinal negative side effects of orlistat (such as oily spotting, flatus with discharge, fecal urgency) cause inconsistent compliance. This greatly reduces the use of this anti-obesity drug for the prevention of diabetes.
Rimonobant, the first selective cannabinoid type 1 (CB1) receptor antagonist, developed as an anti-obesity agent, is nearly completing phase III clinical trials. Originally, it was assumed that a CB1 antagonist would primarily reduce the food intake through a central nervous system-mediated effects [79,80]. Recent emerging evidence in experimental animals clearly indicate that the reduction of body weight and fat mass effected by CB1 antagonists is also due to peripheral mechanisms [81]. These may include a decrease in lipogenesis [82] and an increase in fatty acid oxidation [83]. The data from clinical trials of rimonobant further suggests that the weight loss, derived from the combined anorectic effect and the peripheral healthy effects in modulating fat metabolism, may have benefit for the treatment of Metabolic Syndrome. Rimonabant in Obesity Europe (RIO Europe) is the first of four Phase III studies whose partial data have been revealed [84]. Results of the RIO Europe program indicate that rimonabant at 20 mg/day, is associated with a significant decrease in body weight, as well as with a substantial mobilization of abdominal adipose tissue as indicated by a considerable reduction in waist circumference. Furthermore, the administration of rimonabant produced beneficial metabolic effects on plasma lipid parameters and on the glycemic profile, including the improvement in insulin sensitivity. Thus, these improvements decrease several parameters which comprise the metabolic syndrome in the patient population. As RIO Europe is continuing, it is much anticipated that rimonabant treatment will have significant effect on reducing the incidence of type 2 diabetes among the ~1000 patients with BMI ≥ 30 kg/m2.
Other mechanisms for diabetes prevention
In the original design of the DPP trial, troglitazone, a synthetic agonist of peroxisome proliferator-activated receptor (PPAR)-γ, known for its insulin-sensitizing effect, was included as an independent therapeutic arm [85]. Due to the concern of liver toxicity of troglitazone, this arm was discontinued in 1998 [86]. Using the "short-term" data derived from the mean 0.9 year of troglitazone treatment, the diabetes incidence rate was estimated to be 3.0 cases/100 person-years, compared with 12.0, 6.7, and 5.1 cases/100 person-years in the placebo, metformin, and lifestyle modification participants [86]. The delay of diabetes by troglitazone was associated with preservation of β-cell compensation for insulin resistance [87]. The therapeutic benefit of troglitazone appeared to be reversible. During the 3 years after troglitazone withdrawal, the diabetes incidence rate bounced back to a level similar to the placebo group [86]. According to these data, insulin sensitization through a PPAR-γ activation mechanism is effective in diabetes prevention. However the availability of an agent with an appropriate safety property remains to be a challenge for long term and wide uses.
Acarbose, an α-glucosidase inhibitor, that inhibits the hydrolysis of non-absorbable oligosaccharides and polysaccharides into absorbable monosaccharides which takes place in the brush border of enterocytes, is another agent that is efficacious in reducing blood glucose and insulin level, especially in the postprandial conditions. In the STOP-NIDDM trial, the effect of Acarbose in preventing or delaying the development of type 2 diabetes was assessed [88]. In this multicenter, placebo-controlled randomised trial, patients with IGT were randomly allocated to 100 mg acarbose or placebo three times daily. After a mean follow-up of 3.3 years, 32% patients randomised to acarbose and 42% randomised to placebo developed diabetes. Furthermore, acarbose significantly increased reversion of impaired glucose tolerance to normal glucose tolerance. It was concluded that Acarbose could be used, either as an alternative or in addition to changes in lifestyle, to delay development of type 2 diabetes in patients with IGT.
If preservation of β-cell compensation is an effective way to prevent the progression of type 2 diabetes, strategies based on glucagons-like peptide (GLP) -1 should be successful as well. GLP-1 is an incretin hormone stimulating the glucose-induced insulin secretion in pancreatic β-cells [89]. It not only stimulates the biosynthesis and exocytosis of insulin, but it also has effects on β-cell growth and survival that lead to increased β-cell mass. Moreover, GLP-1 also inhibits the motility in the gastrointestinal tract which slows down the passage of nutrients to the small intestine. When administered in peripheral, GLP-1 also increases the satiety level [90]. In the clinic, a synthetic GLP-1 receptor agonist with a pro-longed half-life compared to the endogenous GLP-1 significantly improved HbA1C [91] along with obvious weight reduction in the patients [92]. Inhibitors of dipeptidyl peptidase (DPP) IV, a protease responsible for the degradation of short-lived endogenous GLP-1, not only extended the half-life of this incretin hormone, but also reduced the excursion of glucose in diabetic patients [93]. These data demonstrate that GLP-1 based therapies, either through administration of exogenous synthetic GLP-1 receptor agonists or through inhibiting DPP IV, are effective in diabetes treatment. Whether these strategies will be successful in pre-diabetes patients to prevent or to delay the onset of type 2 diabetes, awaits more clinical studies designed to address the issue.
Concluding thoughts
Both the diabetes and the obesity epidemics have posed serious assaults on health care. It is anticipated that it will be an uphill battle to fight these ever increasing epidemics, especially since, individually, these conditions are difficult to treat. However, considering the history in the advances of modern medicine, the future looks brighter. The pharmaceutical breakthrough of the statin-class of safe and effective drugs for plasma cholesterol control has caused the mortality rate for cardiovascular disease to decrease [94,95]. The economic benefits have also been demonstrated [96]. Today, there is wide public awareness that obesity and a sedentary lifestyle are culprits leading to the development of diabetes. It is generally known that modifying the sedentary lifestyle with exercise can have a positive impact on diabetes. There are also active efforts in modulating the dietary composition, particularly through cutting down carbohydrates, in the public. What is lacking presently is a true breakthrough for a safe and effective way to treat weight problems. The good news is that there are many appetite suppressant drugs and other anti-obesity drugs in discovery pipelines of the pharmaceutical industry. As more of these future drugs move forward in the pipeline and eventually to the market, it is possible to be optimistic that an era of type 2 diabetes prevention may be about to begin.
Table 3 Potential therapies for the prevention of type 2 diabetes.
Non-pharmaceutical Approach
Method Clinically Proven
Lifestyle modification yes
Bariatric surgery yes
Pharmaceutical Approach
Drugs Molecular Target Site(s) of Action Clinically Proven
Metformin Unknown Liver (muscle) Yes
Acarbose α-glucosidase Intestine Yes
Olistat lipase Intestine Yes
Thiazolidinediones PPARγ Liver, fat, muscle Yes
Rimonobant CB-1 Brain (fat, liver) No
DPP4 inhibitors DPP4 pancreatic islet beta-cells No
GLP1 analogs GLP1-receptor pancreatic islet beta-cells No
Acknowledgements
I am grateful to Simeon Taylor for thoughtful suggestions and to Thomas Harrity for critically reading the manuscript.
==== Refs
Centers for Disease Control and Prevention National diabetes fact sheet: general information and national estimates on diabetes in the United States, 2003. 2003 Acessed at www.cdc.gov/diabetes/pubs/factsheet.htm on June 29th, 2005
Zimmet P Alberti KG Shaw J Global and societal implications of the diabetes epidemic Nature 2001 414 782 787 11742409 10.1038/414782a
Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus Diabetes Care 1997 20 1183 1197 9203460
Joslin EP The prevention of diabetes mellitus JAMA 1921 76 79 84
Kenny SJ Aubert RE Geiss LS Prevalence and incidence of non-insulin-dependent diabetes. Harris MI, Coweie CC, Stern MP et al, eds Diabetes in America National Institute of Health; Washington, DC 1995 47 67
Narayan KM Boyle JP Thompson TJ Sorensen SW Williamson DF Lifetime risk for diabetes mellitus in the United States Jama 2003 290 1884 1890 14532317 10.1001/jama.290.14.1884
Zimmet P Dowse G Finch C Serjeantson S King H The epidemiology and natural history of NIDDM--lessons from the South Pacific Diabetes Metab Rev 1990 6 91 124 2198152
Soderberg S Zimmet P Tuomilehto J de Courten M Dowse GK Chitson P Gareeboo H Alberti KG Shaw JE Increasing prevalence of Type 2 diabetes mellitus in all ethnic groups in Mauritius Diabet Med 2005 22 61 68 15606693 10.1111/j.1464-5491.2005.01366.x
Pan XR Yang WY Li GW Liu J Prevalence of diabetes and its risk factors in China, 1994. National Diabetes Prevention and Control Cooperative Group Diabetes Care 1997 20 1664 1669 9353605
Ramachandran A Snehalatha C Baskar AD Mary S Kumar CK Selvam S Catherine S Vijay V Temporal changes in prevalence of diabetes and impaired glucose tolerance associated with lifestyle transition occurring in the rural population in India Diabetologia 2004 47 860 865 15114469 10.1007/s00125-004-1387-6
Alberti G Zimmet P Shaw J Bloomgarden Z Kaufman F Silink M Type 2 diabetes in the young: the evolving epidemic: the international diabetes federation consensus workshop Diabetes Care 2004 27 1798 1811 15220270
American Diabetes Association Type 2 diabetes in children and adolescents. American Diabetes Association Pediatrics 2000 105 671 680 10699131 10.1542/peds.105.3.671
Songer TJ The economic costs of NIDDM Diabetes Metab Rev 1992 8 389 404 1307526
Engelgau MM Geiss LS Saaddine JB Boyle JP Benjamin SM Gregg EW Tierney EF Rios-Burrows N Mokdad AH Ford ES Imperatore G Narayan KM The evolving diabetes burden in the United States Ann Intern Med 2004 140 945 950 15172919
Mohan V Why are Indians more prone to diabetes? J Assoc Physicians India 2004 52 468 474 15645957
Bennett PH Rewers MJ Knowler WC Epidemiology of diabetes mellitus In : Daniel Porte, Jr and Robert S Sherwin edt Ellenberg and Rifkin's Diabetes Mellitus 1996 fifth edition 373 400
Motala AA Omar MA Pirie FJ Diabetes in Africa. Epidemiology of type 1 and type 2 diabetes in Africa J Cardiovasc Risk 2003 10 77 83 12668904 10.1097/00043798-200304000-00002
Rotimi CN Chen G Adeyemo AA Furbert-Harris P Parish-Gause D Zhou J Berg K Adegoke O Amoah A Owusu S Acheampong J Agyenim-Boateng K Eghan BAJ Oli J Okafor G Ofoegbu E Osotimehin B Abbiyesuku F Johnson T Rufus T Fasanmade O Kittles R Daniel H Chen Y Dunston G Collins FS A genome-wide search for type 2 diabetes susceptibility genes in West Africans: the Africa America Diabetes Mellitus (AADM) Study Diabetes 2004 53 838 841 14988271
Dowse GK Gareeboo H Zimmet PZ Alberti KG Tuomilehto J Fareed D Brissonnette LG Finch CF High prevalence of NIDDM and impaired glucose tolerance in Indian, Creole, and Chinese Mauritians. Mauritius Noncommunicable Disease Study Group Diabetes 1990 39 390 396 2307296
Gohdes D Diabetes in North American Indians and Alaska natives Harris MI, Coweie CC, Stern MP et al, eds Diabetes in America National Institute of Health; Washington, DC 1995 683 702
Flegal KM Carroll MD Ogden CL Johnson CL Prevalence and trends in obesity among US adults, 1999-2000 Jama 2002 288 1723 1727 12365955 10.1001/jama.288.14.1723
Flegal KM Carroll MD Kuczmarski RJ Johnson CL Overweight and obesity in the United States: prevalence and trends, 1960-1994 Int J Obes Relat Metab Disord 1998 22 39 47 9481598 10.1038/sj.ijo.0800541
Allison DB Fontaine KR Manson JE Stevens J VanItallie TB Annual deaths attributable to obesity in the United States Jama 1999 282 1530 1538 10546692 10.1001/jama.282.16.1530
Thorpe K Florence C Howard D The impact of obesity on rising medical spending Health Affairs web exclusive 2004 W4 480 486
Kopelman PG Obesity as a medical problem Nature 2000 404 635 643 10766250
Hu FB Manson JE Stampfer MJ Colditz G Liu S Solomon CG Willett WC Diet, lifestyle, and the risk of type 2 diabetes mellitus in women N Engl J Med 2001 345 790 797 11556298 10.1056/NEJMoa010492
Stein CJ Colditz GA The epidemic of obesity J Clin Endocrinol Metab 2004 89 2522 2525 15181019 10.1210/jc.2004-0288
Basciano H Federico L Adeli K Fructose, insulin resistance, and metabolic dyslipidemia Nutr Metab (Lond) 2005 2 5 15723702 10.1186/1743-7075-2-5
Elliott SS Keim NL Stern JS Teff K Havel PJ Fructose, weight gain, and the insulin resistance syndrome Am J Clin Nutr 2002 76 911 922 12399260
McGarry JD Banting lecture 2001: dysregulation of fatty acid metabolism in the etiology of type 2 diabetes Diabetes 2002 51 7 18 11756317
Shulman GI Unraveling the cellular mechanism of insulin resistance in humans: new insights from magnetic resonance spectroscopy Physiology (Bethesda) 2004 19 183 190 15304632
Yuan M Konstantopoulos N Lee J Hansen L Li ZW Karin M Shoelson SE Reversal of obesity- and diet-induced insulin resistance with salicylates or targeted disruption of Ikkbeta Science 2001 293 1673 1677 11533494 10.1126/science.1061620
Hotamisligil GS Peraldi P Budavari A Ellis R White MF Spiegelman BM IRS-1-mediated inhibition of insulin receptor tyrosine kinase activity in TNF-alpha- and obesity-induced insulin resistance Science 1996 271 665 668 8571133
Shulman GI Cellular mechanisms of insulin resistance J Clin Invest 2000 106 171 176 10903330
Summers SA Nelson DH A role for sphingolipids in producing the common features of type 2 diabetes, metabolic syndrome X, and Cushing's syndrome Diabetes 2005 54 591 602 15734832
Zammit VA Insulin stimulation of hepatic triacylglycerol secretion in the insulin-replete state: implications for the etiology of peripheral insulin resistance Ann N Y Acad Sci 2002 967 52 65 12079835
Shimomura I Matsuda M Hammer RE Bashmakov Y Brown MS Goldstein JL Decreased IRS-2 and increased SREBP-1c lead to mixed insulin resistance and sensitivity in livers of lipodystrophic and ob/ob mice Mol Cell 2000 6 77 86 10949029 10.1016/S1097-2765(00)00009-5
Unger RH Lipotoxicity in the pathogenesis of obesity-dependent NIDDM. Genetic and clinical implications Diabetes 1995 44 863 870 7621989
Kahn BB Flier JS Obesity and insulin resistance J Clin Invest 2000 106 473 481 10953022
Reaven GM Banting lecture 1988. Role of insulin resistance in human disease Diabetes 1988 37 1595 1607 3056758
Reaven GM The Insulin Resistance Syndrome: Definition and Dietary Approaches to Treatment Annu Rev Nutr 2004
Executive Summary of The Third Report of The National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, And Treatment of High Blood Cholesterol In Adults (Adult Treatment Panel III) Jama 2001 285 2486 2497 11368702 10.1001/jama.285.19.2486
World Health Organization Diabetes Mellitus: Report of a WHO Study Group (WHO, Geneva)
Ford ES Giles WH Dietz WH Prevalence of the metabolic syndrome among US adults: findings from the third National Health and Nutrition Examination Survey Jama 2002 287 356 359 11790215 10.1001/jama.287.3.356
Reaven GM The metabolic syndrome: requiescat in pace Clin Chem 2005 51 931 938 15746300 10.1373/clinchem.2005.048611
Grundy SM Point: the metabolic syndrome still lives Clin Chem 2005 51 1352 1354 16040840 10.1373/clinchem.2005.050989
Kahn R Buse J Ferrannini E Stern M The metabolic syndrome: time for a critical appraisal: joint statement from the American Diabetes Association and the European Association for the Study of Diabetes Diabetes Care 2005 28 2289 2304 16123508
American Diabetes Association Expert Committee Report of the Expert Committee Diabetes Care 1997 20 1183 1197 9203460
American Diabetes Association Prediabetes Accessed June 20, 2005
Shaw JE Hodge AM de Courten M Chitson P Zimmet PZ Isolated post-challenge hyperglycaemia confirmed as a risk factor for mortality Diabetologia 1999 42 1050 1054 10447514 10.1007/s001250051269
Tominaga M Eguchi H Manaka H Igarashi K Kato T Sekikawa A Impaired glucose tolerance is a risk factor for cardiovascular disease, but not impaired fasting glucose. The Funagata Diabetes Study Diabetes Care 1999 22 920 924 10372242
Zimmet PZ Alberti KG The changing face of macrovascular disease in non-insulin-dependent diabetes mellitus: an epidemic in progress Lancet 1997 350 Suppl 1 SI1 4 9250275
Benjamin SM Valdez R Geiss LS Rolka DB Narayan KM Estimated number of adults with prediabetes in the US in 2000: opportunities for prevention Diabetes Care 2003 26 645 649 12610015
Barsh GS Farooqi IS O'Rahilly S Genetics of body-weight regulation Nature 2000 404 644 651 10766251
O'Rahilly S Barroso I Wareham NJ Genetic factors in type 2 diabetes: the end of the beginning? Science 2005 307 370 373 15662000 10.1126/science.1104346
Permutt MA Wasson J Cox N Genetic epidemiology of diabetes J Clin Invest 2005 115 1431 1439 15931378 10.1172/JCI24758
Pan XR Li GW Hu YH Wang JX Yang WY An ZX Hu ZX Lin J Xiao JZ Cao HB Liu PA Jiang XG Jiang YY Wang JP Zheng H Zhang H Bennett PH Howard BV Effects of diet and exercise in preventing NIDDM in people with impaired glucose tolerance. The Da Qing IGT and Diabetes Study Diabetes Care 1997 20 537 544 9096977
Tuomilehto J Lindstrom J Eriksson JG Valle TT Hamalainen H Ilanne-Parikka P Keinanen-Kiukaanniemi S Laakso M Louheranta A Rastas M Salminen V Uusitupa M Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance N Engl J Med 2001 344 1343 1350 11333990 10.1056/NEJM200105033441801
Knowler WC Barrett-Connor E Fowler SE Hamman RF Lachin JM Walker EA Nathan DM Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin N Engl J Med 2002 346 393 403 11832527 10.1056/NEJMoa012512
Ramachandran A Simon M Mukesh B Bhaskar D Vijay V Lifestyle modification reduces incident diabetes in Asian Indian with IGT Indian Diabetes Prevention Programme (IDPP1) Diabetes 2005 54 A90
American Diabetes Association and National Institute of Diabetes, Digestive and Kidney Disease The Prevention or Delay of Type 2 Diabetes Diabetes Care 2002 25 742 749 11919135
Zhou G Myers R Li Y Chen Y Shen X Fenyk-Melody J Wu M Ventre J Doebber T Fujii N Musi N Hirshman MF Goodyear LJ Moller DE Role of AMP-activated protein kinase in mechanism of metformin action J Clin Invest 2001 108 1167 1174 11602624 10.1172/JCI200113505
Hardie DG AMP-activated protein kinase: a key system mediating metabolic responses to exercise Med Sci Sports Exerc 2004 36 28 34 14707764
Arora SK Mcfarlane SI The case for low carbohydrate diets in diabetes managerment. Nutr Metab (Lond) 2005 2 16 16018812 10.1186/1743-7075-2-16
Sears B Mastering the zone New York, NY: Harper-Collins Publisher, Inc 1997
Eades ME Eades MD Protein Power New York, NY: Bantam Books 1996
Atkins RC Dr. Atkins' New Diet Revolution Rev ed New York, NY: Avon Books 1998
Foster GD Wyatt HR Hill JO McGuckin BG Brill C Mohammed BS Szapary PO Rader DJ Edman JS Klein A randomized trial of a low-carbohydrate diet for obesity N Engl J Med 2003 348 2082 2090 12761365 10.1056/NEJMoa022207
Samaha FF Iqbal N Seshadri P Chicano KL Daily DA McGrory J Williams M Gracely EJ Stern L A low-carbohydrate as compared with a low-fat diet in severe obesity
N Engl J Med 2003 348 2074 2081 12761364 10.1056/NEJMoa022637
Volek JS Sharman MJ Gomez AL Judelson DA Rubin MR Watson G Sokmen B Silvestre R French DN Kraemer WJ Comparison of energy-restricted very low-carbohydrate and low-fat diets on weight loss and body composition in overweight men and women
Nutr Metab (Lond) 2004 1 13 15533250 10.1186/1743-7075-1-13
Boden G Sargrad K Homko C Mozzoli M Stein TP Effect of a low-carbohydrate diet on appetite, blood glucose levels, and insulin resistance in obese patients with type 2 diabetes Ann Intern Med 2005 142 403 411 15767618
Gannon MC Nuttall FQ Effect of a high-protein, low-carbohydrate diet on blood glucose control in people with type 2 diabetes Diabetes 2004 53 2375 2382 15331548
Long SD O'Brien K MacDonald KGJ Leggett-Frazier N Swanson MS Pories WJ Caro JF Weight loss in severely obese subjects prevents the progression of impaired glucose tolerance to type II diabetes. A longitudinal interventional study Diabetes Care 1994 17 372 375 8062602
Sjostrom CD Peltonen M Wedel H Sjostrom L Differentiated long-term effects of intentional weight loss on diabetes and hypertension Hypertension 2000 36 20 25 10904007
Goldstein DJ Beneficial health effects of modest weight loss Int J Obes Relat Metab Disord 1992 16 397 415 1322866
Rissanen A Pharmacological intervention: the antiobesity approach Eur J Clin Invest 1998 28 Suppl 2 27 30 9777325 10.1046/j.1365-2362.1998.0280s2027.x
Torgerson JS Hauptman J Boldrin MN Sjostrom L XENical in the prevention of diabetes in obese subjects (XENDOS) study: a randomized study of orlistat as an adjunct to lifestyle changes for the prevention of type 2 diabetes in obese patients Diabetes Care 2004 27 155 161 14693982
Hollander PA Elbein SC Hirsch IB Kelley D McGill J Taylor T Weiss SR Crockett SE Kaplan RA Comstock J Lucas CP Lodewick PA Canovatchel W Chung J Hauptman J Role of orlistat in the treatment of obese patients with type 2 diabetes. A 1-year randomized double-blind study Diabetes Care 1998 21 1288 1294 9702435
Simiand J Keane M Keane PE Soubrie P SR 141716, a CB1 cannabinoid receptor antagonist, selectively reduces sweet food intake in marmoset Behav Pharmacol 1998 9 179 181 10065938
Colombo G Agabio R Diaz G Lobina C Reali R Gessa GL Appetite suppression and weight loss after the cannabinoid antagonist SR 141716 Life Sci 1998 63 PL113 7 9718088 10.1016/S0024-3205(98)00322-1
Vickers SP Webster LJ Wyatt A Dourish CT Kennett GA Preferential effects of the cannabinoid CB1 receptor antagonist, SR 141716, on food intake and body weight gain of obese (fa/fa) compared to lean Zucker rats Psychopharmacology (Berl) 2003 167 103 111 12632249
Osei-Hyiaman D DePetrillo M Pacher P Liu J Radaeva S Batkai S Harvey-White J Mackie K Offertaler L Wang L Kunos G Endocannabinoid activation at hepatic CB1 receptors stimulates fatty acid synthesis and contributes to diet-induced obesity J Clin Invest 2005 115 1298 1305 15864349 10.1172/JCI200523057
Bensaid M Gary-Bobo M Esclangon A Maffrand JP Le Fur G Oury-Donat F Soubrie P The cannabinoid CB1 receptor antagonist SR141716 increases Acrp30 mRNA expression in adipose tissue of obese fa/fa rats and in cultured adipocyte cells Mol Pharmacol 2003 63 908 914 12644592 10.1124/mol.63.4.908
Van Gaal LF Rissanen AM Scheen AJ Ziegler O Rossner S Effects of the cannabinoid-1 receptor blocker rimonabant on weight reduction and cardiovascular risk factors in overweight patients: 1-year experience from the RIO-Europe study Lancet 2005 365 1389 1397 15836887 10.1016/S0140-6736(05)66374-X
The Diabetes Prevention Program. Design and methods for a clinical trial in the prevention of type 2 diabetes Diabetes Care 1999 22 623 634 10189543
Knowler WC Hamman RF Edelstein SL Barrett-Connor E Ehrmann DA Walker EA Fowler SE Nathan DM Kahn SE Prevention of type 2 diabetes with troglitazone in the Diabetes Prevention Program Diabetes 2005 54 1150 1156 15793255
Buchanan TA Xiang AH Peters RK Kjos SL Marroquin A Goico J Ochoa C Tan S Berkowitz K Hodis HN Azen SP Preservation of pancreatic beta-cell function and prevention of type 2 diabetes by pharmacological treatment of insulin resistance in high-risk hispanic women Diabetes 2002 51 2796 2803 12196473
Chiasson JL Josse RG Gomis R Hanefeld M Karasik A Laakso M Acarbose for prevention of type 2 diabetes mellitus: the STOP-NIDDM randomised trial Lancet 2002 359 2072 2077 12086760 10.1016/S0140-6736(02)08905-5
Mojsov S Weir GC Habener JF Insulinotropin: glucagon-like peptide I (7-37) co-encoded in the glucagon gene is a potent stimulator of insulin release in the perfused rat pancreas J Clin Invest 1987 79 616 619 3543057
Deacon CF Therapeutic strategies based on glucagon-like peptide 1 Diabetes 2004 53 2181 2189 15331525
Egan JM Meneilly GS Elahi D Effects of 1-mo bolus subcutaneous administration of exendin-4 in type 2 diabetes Am J Physiol Endocrinol Metab 2003 284 E1072 9 12475750
Baron A Poon T Taylor K Nielsen L Boies S Zhou J Zhuang D Varns A Kim D Fineman M Kolterman O Exenadite (synthetic extendin-4) showed marked HbA1c decline over 5 months in patients with type 2 diabetes failing oral agents in an open-label study Presented at the 63rd Scientific Sessions of the American Diabetes Association, New Orleans, LA, 13-17 June 2003 (late breaking abstract 3-LB) 2003
Ahren B Simonsson E Larsson H Landin-Olsson M Torgeirsson H Jansson PA Sandqvist M Bavenholm P Efendic S Eriksson JW Dickinson S Holmes D Inhibition of dipeptidyl peptidase IV improves metabolic control over a 4-week study period in type 2 diabetes Diabetes Care 2002 25 869 875 11978683
Tyroler HA Cholesterol and cardiovascular disease. An overview of Lipid Research Clinics (LRC) epidemiologic studies as background for the LRC Coronary Primary Prevention Trial
Am J Cardiol 1984 54 14C 19C 6741804 10.1001/jama.278.4.313
Hebert PR Gaziano JM Chan KS Hennekens CH Cholesterol lowering with statin drugs, risk of stroke, and total mortality. An overview of randomized trials Jama 1997 278 313 321 9228438 10.1001/jama.278.4.313
Pharoah PD Hollingworth W Cost effectiveness of lowering cholesterol concentration with statins in patients with and without pre-existing coronary heart disease: life table method applied to health authority population Bmj 1996 312 1443 1448 8664620
Inoue H Ferrer J Welling CM Elbein SC Hoffman M Mayorga R Warren-Perry M Zhang Y Millns H Turner R Province M Bryan J Permutt MA Aguilar-Bryan L Sequence variants in the sulfonylurea receptor (SUR) gene are associated with NIDDM in Caucasians Diabetes 1996 45 825 831 8635661
Hani EH Clement K Velho G Vionnet N Hager J Philippi A Dina C Inoue H Permutt MA Basdevant A North M Demenais F Guy-Grand B Froguel P Genetic studies of the sulfonylurea receptor gene locus in NIDDM and in morbid obesity among French Caucasians Diabetes 1997 46 688 694 9075812
Horikawa Y Oda N Cox NJ Li X Orho-Melander M Hara M Hinokio Y Lindner TH Mashima H Schwarz PE del Bosque-Plata L Oda Y Yoshiuchi I Colilla S Polonsky KS Wei S Concannon P Iwasaki N Schulze J Baier LJ Bogardus C Groop L Boerwinkle E Hanis CL Bell GI Genetic variation in the gene encoding calpain-10 is associated with type 2 diabetes mellitus Nat Genet 2000 26 163 175 11017071 10.1038/79876
Weedon MN Schwarz PE Horikawa Y Iwasaki N Illig T Holle R Rathmann W Selisko T Schulze J Owen KR Evans J Del Bosque-Plata L Hitman G Walker M Levy JC Sampson M Bell GI McCarthy MI Hattersley AT Frayling TM Meta-analysis and a large association study confirm a role for calpain-10 variation in type 2 diabetes susceptibility Am J Hum Genet 2003 73 1208 1212 14574648 10.1086/379285
Hager J Hansen L Vaisse C Vionnet N Philippi A Poller W Velho G Carcassi C Contu L Julier C Cambien F Passa P Lathrop M Kindsvogel W Demenais F Nishimura E Froguel P A missense mutation in the glucagon receptor gene is associated with non-insulin-dependent diabetes mellitus Nature Genetics 9 1995 9 299 394 10.1038/ng0395-299
McCarthy MI Hitman GA Hitchins M Riikonen A Stengard J Nissinen A Tuomilehto-Wolf E Tuomilehto J Glucokinase gene polymorphisms: a genetic marker for glucose intolerance in a cohort of elderly Finnish men Diabet Med 1994 11 198 204 8200207
Hani EH Boutin P Durand E Inoue H Permutt MA Velho G Froguel P Missense mutations in the pancreatic islet beta cell inwardly rectifying K+ channel gene (KIR6.2/BIR): a meta-analysis suggests a role in the polygenic basis of Type II diabetes mellitus in Caucasians Diabetologia 1998 41 1511 1515 9867219 10.1007/s001250051098
Deeb SS Fajas L Nemoto M Pihlajamaki J Mykkanen L Kuusisto J Laakso M Fujimoto W Auwerx J A Pro12Ala substitution in PPARgamma2 associated with decreased receptor activity, lower body mass index and improved insulin sensitivity Nat Genet 1998 20 284 287 9806549 10.1038/3099
Yamagata K Furuta H Oda N Kaisaki PJ Menzel S Cox NJ Fajans SS Signorini S Stoffel M Bell GI Mutations in the hepatocyte nuclear factor-4alpha gene in maturity-onset diabetes of the young (MODY1) Nature 1996 384 458 460 8945471 10.1038/384458a0
Yamagata K Oda N Kaisaki PJ Menzel S Furuta H Vaxillaire M Southam L Cox RD Lathrop GM Boriraj VV Chen X Cox NJ Oda Y Yano H Le Beau MM Yamada S Nishigori H Takeda J Fajans SS Hattersley AT Iwasaki N Hansen T Pedersen O Polonsky KS Bell GI Mutations in the hepatocyte nuclear factor-1alpha gene in maturity-onset diabetes of the young (MODY3) Nature 1996 384 455 458 8945470 10.1038/384455a0
Li SR Baroni MG Oelbaum RS Stock J Galton DJ Association of genetic variant of the glucose transporter with non-insulin-dependent diabetes mellitus Lancet 1988 2 368 370 2899775 10.1016/S0140-6736(88)92836-X
Huxtable SJ Saker PJ Haddad L Walker M Frayling TM Levy JC Hitman GA O'Rahilly S Hattersley AT McCarthy MI Analysis of parent-offspring trios provides evidence for linkage and association between the insulin gene and type 2 diabetes mediated exclusively through paternally transmitted class III variable number tandem repeat alleles Diabetes 2000 49 126 130 10615960
Kadowaki T Bevins CL Cama A Ojamaa K Marcus-Samuels B Kadowaki H Beitz L McKeon C Taylor SI Two mutant alleles of the insulin receptor gene in a patient with extreme insulin resistance Science 1988 240 787 790 2834824
Taylor SI Kadowaki T Kadowaki H Accili D Cama A McKeon C Mutations in insulin-receptor gene in insulin-resistant patients Diabetes Care 1990 13 257 279 1968373
Poulton J Luan J Macaulay V Hennings S Mitchell J Wareham NJ Type 2 diabetes is associated with a common mitochondrial variant: evidence from a population-based case-control study Hum Mol Genet 2002 11 1581 1583 12045211 10.1093/hmg/11.13.1581
Clement K Vaisse C Lahlou N Cabrol S Pelloux V Cassuto D Gourmelen M Dina C Chambaz J Lacorte JM Basdevant A Bougneres P Lebouc Y Froguel P Guy-Grand B A mutation in the human leptin receptor gene causes obesity and pituitary dysfunction Nature 1998 392 398 401 9537324 10.1038/32911
Yeo GS Farooqi IS Aminian S Halsall DJ Stanhope RG O'Rahilly S A frameshift mutation in MC4R associated with dominantly inherited human obesity Nat Genet 1998 20 111 112 9771698 10.1038/2404
Montague CT Farooqi IS Whitehead JP Soos MA Rau H Wareham NJ Sewter CP Digby JE Mohammed SN Hurst JA Cheetham CH Earley AR Barnett AH Prins JB O'Rahilly S Congenital leptin deficiency is associated with severe early-onset obesity in humans Nature 1997 387 903 908 9202122 10.1038/43185
Jackson RS Creemers JW Ohagi S Raffin-Sanson ML Sanders L Montague CT Hutton JC O'Rahilly S Obesity and impaired prohormone processing associated with mutations in the human prohormone convertase 1 gene Nat Genet 1997 16 303 306 9207799 10.1038/ng0797-303
Krude H Biebermann H Luck W Horn R Brabant G Gruters A Severe early-onset obesity, adrenal insufficiency and red hair pigmentation caused by POMC mutations in humans Nat Genet 1998 19 155 157 9620771 10.1038/509
Suviolahti E Oksanen LJ Ohman M Cantor RM Ridderstrale M Tuomi T Kaprio J Rissanen A Mustajoki P Jousilahti P Vartiainen E Silander K Kilpikari R Salomaa V Groop L Kontula K Peltonen L Pajukanta P The SLC6A14 gene shows evidence of association with obesity J Clin Invest 2003 112 1762 1772 14660752 10.1172/JCI200317491
|
16232315
|
PMC1309619
|
CC BY
|
2021-01-04 16:37:46
|
no
|
Nutr Metab (Lond). 2005 Oct 18; 2:29
|
utf-8
|
Nutr Metab (Lond)
| 2,005 |
10.1186/1743-7075-2-29
|
oa_comm
|
==== Front
Reprod Biol EndocrinolReproductive biology and endocrinology : RB&E1477-7827BioMed Central London 1477-7827-3-621628007310.1186/1477-7827-3-62ResearchSubcutaneously administered Menopur(R), a new highly purified human menopausal gonadotropin, causes significantly fewer injection site reactions than Repronex(R) in subjects undergoing in vitro fertilization Keye William R [email protected] Bobby [email protected] Richard [email protected] Stephen [email protected] Jack [email protected] M Joseph [email protected] William Beaumont Hospital, In Vitro Fertility Clinic, Royal Oak, Michigan, USA2 Woman's Center for Fertility, Baton Rouge, Louisiana, USA3 Fertility Institute of New Orleans, New Orleans, Louisiana, USA4 Abington Reproductive Medicine, Abington, Pennsylvania, USA5 Reproductive Endocrine Associates of Charlotte, Charlotte, North Carolina, USA6 Ferring Pharmaceuticals Inc., Suffern, New York, USA2005 9 11 2005 3 62 62 27 6 2005 9 11 2005 Copyright © 2005 Keye et al; licensee BioMed Central Ltd.2005Keye et al; licensee BioMed Central Ltd.This is an Open Access article distributed under the terms of the Creative Commons Attribution License (), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Background
The safety and tolerability of a new highly purified, urine-derived human menopausal gonadotropin (hMG) preparation [Menopur(R)] was compared with a currently available hMG [Repronex (R)] in women undergoing in vitro fertilization (IVF).
Methods
This was a randomized, open-label, parallel-group, multicenter study conducted in subjects undergoing IVF. Women (N = 125), 18–39 years of age, underwent pituitary down-regulation with leuprolide acetate beginning 7 days prior to onset of menses and continuing up to the day before hCG administration. Subjects were randomized to receive subcutaneous (SC) Menopur (R) (n = 61) or Repronex (R) SC (n = 64) for a maximum of 12 days. All adverse events (AEs) were recorded and subject self-assessments of injection site reactions were recorded in a daily diary.
Results
Significantly fewer subjects in the Menopur (R) group reported injection site reactions (P < 0.001) compared to the Repronex (R) group. Overall, there was no statistically significant difference in the incidence of AEs between the two treatment groups.
Conclusion
Menopur (R) SC offers a greater safety and tolerability profile compared to Repronex (R) SC.
==== Body
Background
Zondek and colleagues were the first to propose that the pituitary gland secretes hormones that stimulate the gonads [1]. This hypothesis was later confirmed with the identification of two different hormones, follicle-stimulating hormone (FSH) and luteinizing hormone (LH) [2]. These advances in the understanding of the human reproductive process converged in 1958 with the successful clinical use of pituitary gonadotropins to induce ovulation in anovulatory women [3].
Development of simple extraction and purification techniques led to the production of human menopausal gonadotropins (hMG) in quantities sufficient for clinical use. In 1962, the first pregnancy resulting from the use of a urine-derived hMG for follicular stimulation was reported [4]. Since then, human-derived gonadotropins have remained a reliable and safe treatment for infertility. However, the purity of early hMG preparations was low and the majority of the injectant preparations consisted of uncharacterized urinary proteins [5]. The uncharacterized proteins produced adverse injection site reactions when administered intramuscularly (IM) and the product could not be administered subcutaneously (SC). Newer purification techniques applied to the manufacturing of hMG resulted in enhanced purity and enabled SC administration. However, mild to moderate injection site reactions were still common. Most recently, modern day purification techniques have resulted in the availability of a new, high purity hMG preparation that has nearly eliminated injection site reactions.
The purification process for Repronex® included at least 24 steps involving adsorptions, dialysis and precipitations, as well as ionic, cationic, and hydrophobic exchange chromatography [6]. Advances in these manufacturing techniques and the inclusion of additional purification steps have now produced a new highly purified hMG (Menopur® , 75 IU LH:75 IU FSH; Ferring Pharmaceuticals Inc., Suffern, New York) that is nearly devoid of uncharacterized proteins. This high purity hMG was developed to reduce the frequency of injection site reactions. Therefore, we conducted a clinical trial to compare the safety and tolerability of Menopur® with the currently available hMG preparation, Repronex® (75 IU LH:75 IU FSH; Ferring Pharmaceuticals Inc.) in women undergoing controlled ovarian hyperstimulation (COH) for in vitro fertilization (IVF).
Methods
This was a randomized, open-label, parallel-group, multicenter study comparing one cycle of treatment with Menopur® SC or Repronex® SC at a dose of 75 to 450 IU/day administered as a single, daily injection for up to 12 days in infertile women undergoing COH for IVF. Each subject participated in only one cycle of IVF. Fifteen centers participated in the study, each of which obtained institutional review board approval. Written informed consent was obtained from all participants prior to screening and study enrollment.
Subjects
Subjects had to meet the following eligibility criteria to participate in the study: nonsmoking, 18 to 39 years of age with regular ovulatory cycles (24 to 35 days); have a diagnosis of unexplained infertility or infertility due to tubal factor; stage I or II endometriosis; normal ovaries and uterus on transvaginal ultrasonography; normal serum levels of estradiol (E2), prolactin, LH, FSH, testosterone, dehydroepiandrosterone sulfate, and thyroid-stimulating hormone; and have a body mass index of ≤ 34. In addition, all subjects were seronegative for hepatitis B and C and HIV and had a negative pregnancy test prior to initiating treatment. A semen analysis performed on a sample from either the subject's partner or the designated donor had to be normal according to the criteria established by the World Health Organization. A minimum of one menstrual cycle without IVF/assisted reproductive therapy (ART) treatment was required prior to screening. Subjects were excluded from participation if there was evidence of any clinically relevant systemic disease or any surgical or medical condition that could interfere with the absorption, metabolism, or excretion of gonadotropins. Subjects were not to have had a positive pregnancy test within three months of baseline screening and were also excluded from the study if they had undergone three or more prior ART cycles, had abnormal uterine bleeding, a history of substance abuse, a history of chemotherapy, were breast feeding, or if they had participated in any experimental drug study within 60 days of screening for this study.
Protocol
Each eligible subject received daily injections of leuprolide acetate (LA; TAP Pharmaceuticals, Deerfield, IL; 0.5 mg/d SC) beginning 7 days prior to the anticipated onset of menses until the day before administration of human chorionic gonadotropin (hCG)(Novarel™, Ferring Pharmaceuticals Inc., Suffern, NY). LA was continued until there was evidence of down-regulation as indicated by a serum E2 concentration of ≤ 45 pg/mL and an endometrial lining ≤ 7 mm on transvaginal ultrasound. If the E2 level was ≥ 45 pg/ml, the endometrial lining was > 7 mm, or menses did not occur within 20 days after beginning LA, the subject was withdrawn from the study.
Subjects who met the down-regulation criteria listed above were randomized to receive single, daily doses of Menopur® SC or Repronex® SC. They were instructed to self-administer hMG SC, alternating their injections between the right and left lower abdomen at approximately the same time every afternoon. Subjects were also instructed to record injection site pain each day using a 0- to 10-point scale, with 0 representing no pain and 10 representing extreme pain. In addition, subjects recorded all adverse events (AEs) experienced during the study, including those associated with injection site reactions. Subjects received 225 IU for 4 days and on day 5, subjects returned to the study centers for transvaginal ultrasound and determination of E2 levels. Based on these findings, the investigators adjusted the daily dose of hMG by 75 to 150 IU up to a maximum of 450 IU per day. Serum E2 levels and ultrasound measurements were taken prior to each dose escalation and a maximum of 12 days total hMG treatment was allowed. Investigators were permitted to decrease the hMG dose at any time based on clinical judgment and safety concerns, and could discontinue hMG and/or withhold hCG administration if they believed the subject was at risk for development of ovarian hyperstimulation syndrome (OHSS). When at least 3 follicles reached a diameter of ≥ 16 mm as measured by transvaginal ultrasound and E2 levels were appropriate for the number of follicles observed based on the investigator's clinical judgment, hMG was discontinued and hCG (Novarel™, Ferring Pharmaceuticals, Inc.) was administered IM at a dose of 10,000 USP units. Oocytes were retrieved 34 to 36 hours later. Standard center-specific IVF culture conditions were allowed, however intracytoplasmic sperm injection (ICSI) and assisted hatching were not. Study centers were permitted to use coculture with homologous cells if it was standard practice and routinely used in all subjects undergoing IVF at that center. A maximum of four embryos could be transferred. Progesterone (Crinone™ 8% gel, 90 mg qd, Serono Laboratories, Inc., Randolph, MA) was self-administered beginning on day 2 or day 3 after oocyte retrieval for luteal phase support, and continued until there was fetal heart motion in an intrauterine pregnancy or there was a negative serum pregnancy test (β-hCG).
Throughout the study, investigators recorded the presence and nature of any AEs. Within 3 weeks of the initial β-hCG quantitative serum pregnancy test or discontinuation from study, subjects returned to the study center for an exit physical examination. Subject reports of AEs were recorded on the case report form and tabulated using Coding Symbols for Thesaurus of Adverse Reaction Terms (COSTART) terminology. Subjects' daily diary evaluations of injection site pain were tabulated on a 0- to 10-point scale.
Statistical evaluation
The primary efficacy of this study was oocytes retrieved and therefore power calculations were done with 80% power to detect an among group difference of 30% in the number of oocytes retrieved. This required 59 patients per group.
A chi-square test was used to make between group comparisons on the percentage of subjects with at least one AE and at least one mild to moderate AE while a Fisher exact test was used to make comparisons on the percentage of subjects with at least one severe AE and at least one serious AE. A chi-square test was used to test for differences in the percentage of subjects with abnormal findings recorded at the study exit physical examination, while the Fisher exact test was used to test for differences in the percentage of subjects with clinically significant abnormal findings. The initial analysis of injection site pain on each treatment day compared the two treatment groups using a one-way ANOVA. A linear mixed model was then used to make treatment comparisons of injection site pain throughout the study. This model allowed the analysis of continuous correlated data to account for within subject variation.
Results
A total of 190 subjects were randomized and included in the analysis of safety. The initial study contained a third arm consisting of 65 subjects who received Menopur® IM, however these data were not included in this analysis, as the focus of this report is to compare the safety and tolerability of Repronex® SC and Menopur® SC. The remaining 125 subjects were randomized to receive Menopur® SC (n = 61) or Repronex® SC (n = 64). Due to subject noncompliance or loss at follow-up, certain safety outcomes such as exit physical examination variables and injection site pain are missing a few data points.
Subject demographic characteristics are summarized in Table 1. Overall, subjects in the two treatment groups were comparable both demographically and medically. The only statistically significant difference between the groups was race, with African-Americans comprising 11.5% of the Menopur® group compared with 1.6% of the Repronex® group (P = 0.039). The impact of this difference is unknown.
Table 1 Demographic Characteristics of Subjects*
Characteristic Menopur® (n = 61) Repronex® (n = 64)
Age (yrs) 32.3 (3.7) 32.5 (4.1)
Weight (kg) 63.5 (11.0) 63.5 (10.0)
Height (cm) 161.5 (7.4) 163.3 (6.4)
Body mass index (kg/m2 ) 24.4 (3.6) 24.0 (3.4)
Race, no. (%)
Caucasian 47 (77.0) 54 (84.4)
African-American 7 (11.5) 1 (1.6)
Asian 0 (0) 2 (3.1)
Hispanic 5 (8.2) 5 (7.8)
Native American 0 (0) 0 (0)
Other 2 (3.3) 2 (3.1)
*Values are mean (SD).
There were no statistically significant differences between the treatment groups in the number of subjects with any AEs, severe AEs, or serious AEs, as shown in Table 2. There were five serious AEs during the study (1 subject in the Menopur® group had OHSS and four subjects in the Repronex® group had one of the following serious AEs: dehydration, an ectopic pregnancy, a right ruptured ovary with secondary hemothorax, and a pelvic abscess). A total of three cases of OHSS were reported (1 subject in the Menopur® group, which was severe and 2 subjects in the Repronex® group, which were mild or moderate).
Table 2 Subjects with Adverse Events*
Adverse Event Menopur® (n = 61) Repronex® (n = 64) P Value
Any 41 (67.2) 48 (75.0) 0.620
Severe 5 (8.2) 5 (7.8) 0.402
Serious 1 (1.6) 4 (6.3) 0.456
*Values represent numbers (percentage) of subjects with one or more adverse event.
Table 3 lists the AEs with an incidence of ≥ 5% (2 or more subjects). Among these AEs, there were no significant differences between the two groups in the percentage of subjects with any AE and no difference in the intensity of injection site pain. However, there were numerically fewer total AEs in the Menopur® group (n = 131) compared to the Repronex® group (n = 198). As shown in Figure 1, this difference was largely attributed to the number of injection site reactions, the single most common AE. When only hMG injections were considered, there were only three (4.9%) subjects in the Menopur® group that reported injection site reactions, whereas 22 (34.4%) subjects in the Repronex® group reported injection site reactions (P < 0.001). Among the three Menopur® subjects with local injection site reactions, all were transient and mild to moderate in intensity, none developed welts/inflammation, and only one subject had localized swelling. These findings contrasted with the 22 subjects in the Repronex® group with injection site reactions, among whom eight developed welts/inflammation (P < 0.001) and four developed swelling (P = 0.328). Conversely, there was no difference in mean scores for injection site pain between the two groups, 2.6 for Menopur® and 2.3 for Repronex® (P = 0.615).
Table 3 Subjects with Adverse Events: Incidence Rate ≥ 5%* (two or more subjects)
Adverse Event Menopur® n = 61 Repronex® n = 64
Abdominal cramps 13 (21.3) 14 (21.9)
Headache 13 (21.3) 13 (20.3)
Post retrieval pain 7 (11.5) 9 (14.1)
Nausea 6 (9.8) 10 (15.6)
Vaginal spotting 6 (9.8) 5 (7.8)
Abdominal fullness 5 (8.2) 7 (10.9)
Abdominal pain 5 (8.2) 4 (6.3)
Constipation 5 (8.2) 1 (1.6)
Respiratory disorder 4 (6.6) 4 (6.3)
Vaginal hemorrhage 2 (3.3) 4 (6.3)
Breast tenderness/pain 2 (3.3) 5 (7.8)
Malaise 2 (3.3) 4 (6.3)
Sinusitis 2 (3.3) 4 (6.3)
*Values represent numbers (percentage) of subjects with adverse event.
Figure 1 Subjects with hMG-Associated Injection Site Reactions. This figure shows the percentage of subjects with any hMG associated injection site reaction as well as those with reactions that included welts or inflammation and those whose reactions involved swelling.
Discussion
Overall, the safety profile of Menopur® in this study was similar to that of Repronex® . Human-derived gonadotropins have been used safely and effectively in ART protocols for over forty years. However, the injection of partially purified hMG is associated with more injection site reactions than highly purified gonadotropins. Removal of nearly all uncharacterized proteins from hMG in the manufacturing process for Menopur® has resulted in significantly fewer reported injection site reactions in IVF subjects. There was a seven-fold difference in the percentage of subjects with injection site reactions, 4.9% and 34.4% of subjects in the Menopur® and Repronex® groups, respectively. When the incidence of reactions that involved swelling, inflammation, or welts was examined, 98% of subjects receiving Menopur® completed their cycle without such reactions while only 81% of subjects receiving Repronex® did not experience such events (P = 0.001).
An analysis of Menopur® has shown that its purity and quality is comparable to recombinant gonadotropin preparations [7]. In addition, Menopur® has been shown to have a similar safety and tolerability profile as recombinant FSH in women undergoing IVF/ICSI treatment cycles [8]. Collectively, these observations and studies, combined with the data from this study demonstrate that Menopur® is at least as efficacious and safe as any existing gonadotropin.
The results from this study demonstrate that Menopur® , a new highly purified hMG, can be administered SC with significantly fewer injection site reactions than Repronex® , a partially purified hMG. Thus, advanced manufacturing techniques have produced the first ever highly purified form of hMG resulting in a markedly improved safety and tolerability profile compared with previously available hMG products.
Authors' contributions
Drs. Keye, Webster, Dickey, Somkuti, and Crain contributed to the treatment of subjects, collection of data and writing of the manuscript. Dr. Scobey was instrumental in data analysis and writing of the manuscript.
==== Refs
Zondek B Weitere untersuchungen zur darstellung, biologie und klinik des hypophysenvorderlappenhormons (Prolan) Zentralbl Gynako 1929 53 834 847
Fevold HL Hisaw FL Leonard SL The gonad stimulating and the luteinizing hormones of the anterior lobe of the hypophysis Am J Physiol 1931 97 291 301
Gemzell CA Diczfalusy E Tillinger G Clinical effect of human pituitary follicle-stimulating hormone (FSH) J Clin Endocrinol Metab 1958 18 1333 1348 13611018
Lunenfeld B Sulimovici S Rabau E Eshkol A L'induction de l'ovulation dans les amenorrhees hypophysaires par un traitement combine de gonadotrophines urinaires menopausiques et de gonadotrophines chorioniques CR Soc Fr Gyncol 1962 32 347 351
Lunenfeld B Historical perspectives in gonadotrophin therapy Hum Reprod Updat 2004 10 453 467 10.1093/humupd/dmh044
Reichl H Balen A Jansen CAM Prion transmission in blood and urine: what are the implications for recombinant and urinary-derived gonadotrophins? Hum Reprod 2002 17 2501 2508 12351519 10.1093/humrep/17.10.2501
Wolfenson C Groisman J Couto AS Hedenfalk M Cortvrindt RG Smitz JE Jesperson S Batch-to-batch consistency of human-derived gonadotrophin preparations compared with recombinant preparations Reprod Biomed Online 2005 10 442 454 15901450
European and Israeli Study Group Efficacy and safety of highly purified menotropin versus recombinant follicle-stimulating hormone in in vitro fertilization/intracytoplasmic sperm injection cycles: a randomized, comparative trial Fertil Steri 2002 78 520 528 10.1016/S0015-0282(02)03250-8
|
16280073
|
PMC1309620
|
CC BY
|
2021-01-04 16:37:13
|
no
|
Reprod Biol Endocrinol. 2005 Nov 9; 3:62
|
utf-8
|
Reprod Biol Endocrinol
| 2,005 |
10.1186/1477-7827-3-62
|
oa_comm
|
==== Front
PLoS Comput BiolPLoS Comput. BiolpcbiplcbploscompPLoS Computational Biology1553-734X1553-7358Public Library of Science San Francisco, USA 1647732410.1371/journal.pcbi.001006705-PLCB-RA-0179R2plcb-01-07-03Research ArticleBioinformatics - Computational BiologyGenetics/Comparative GenomicsSaccharomycesPhyloGibbs: A Gibbs Sampling Motif Finder That Incorporates Phylogeny PhyloGibbs: A Phylogenetic Motif FinderSiddharthan Rahul 12Siggia Eric D 1van Nimwegen Erik 13*1 Center for Studies in Physics and Biology, The Rockefeller University, New York, New York, United States of America
2 Institute of Mathematical Sciences, Taramani, Chennai, India
3 Division of Bioinformatics, Biozentrum, University of Basel, Basel, Switzerland
Bourne Philip EditorUniversity of California at San Diego, United States of America* To whom correspondence should be addressed. E-mail: [email protected] 2005 9 12 2005 27 10 2005 1 7 e6727 7 2005 28 10 2005 Copyright: © 2005 Siddharthan et al.2005This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly credited.A central problem in the bioinformatics of gene regulation is to find the binding sites for regulatory proteins. One of the most promising approaches toward identifying these short and fuzzy sequence patterns is the comparative analysis of orthologous intergenic regions of related species. This analysis is complicated by various factors. First, one needs to take the phylogenetic relationship between the species into account in order to distinguish conservation that is due to the occurrence of functional sites from spurious conservation that is due to evolutionary proximity. Second, one has to deal with the complexities of multiple alignments of orthologous intergenic regions, and one has to consider the possibility that functional sites may occur outside of conserved segments. Here we present a new motif sampling algorithm, PhyloGibbs, that runs on arbitrary collections of multiple local sequence alignments of orthologous sequences. The algorithm searches over all ways in which an arbitrary number of binding sites for an arbitrary number of transcription factors (TFs) can be assigned to the multiple sequence alignments. These binding site configurations are scored by a Bayesian probabilistic model that treats aligned sequences by a model for the evolution of binding sites and “background” intergenic DNA. This model takes the phylogenetic relationship between the species in the alignment explicitly into account. The algorithm uses simulated annealing and Monte Carlo Markov-chain sampling to rigorously assign posterior probabilities to all the binding sites that it reports. In tests on synthetic data and real data from five Saccharomyces species our algorithm performs significantly better than four other motif-finding algorithms, including algorithms that also take phylogeny into account. Our results also show that, in contrast to the other algorithms, PhyloGibbs can make realistic estimates of the reliability of its predictions. Our tests suggest that, running on the five-species multiple alignment of a single gene's upstream region, PhyloGibbs on average recovers over 50% of all binding sites in S. cerevisiae at a specificity of about 50%, and 33% of all binding sites at a specificity of about 85%. We also tested PhyloGibbs on collections of multiple alignments of intergenic regions that were recently annotated, based on ChIP-on-chip data, to contain binding sites for the same TF. We compared PhyloGibbs's results with the previous analysis of these data using six other motif-finding algorithms. For 16 of 21 TFs for which all other motif-finding methods failed to find a significant motif, PhyloGibbs did recover a motif that matches the literature consensus. In 11 cases where there was disagreement in the results we compiled lists of known target genes from the literature, and found that running PhyloGibbs on their regulatory regions yielded a binding motif matching the literature consensus in all but one of the cases. Interestingly, these literature gene lists had little overlap with the targets annotated based on the ChIP-on-chip data. The PhyloGibbs code can be downloaded from http://www.biozentrum.unibas.ch/~nimwegen/cgi-bin/phylogibbs.cgi or http://www.imsc.res.in/~rsidd/phylogibbs. The full set of predicted sites from our tests on yeast are available at http://www.swissregulon.unibas.ch.
Synopsis
Computational discovery of regulatory sites in intergenic DNA is one of the central problems in bioinformatics. Up until recently motif finders would typically take one of the following two general approaches. Given a known set of co-regulated genes, one searches their promoter regions for significantly overrepresented sequence motifs. Alternatively, in a “phylogenetic footprinting” approach one searches multiple alignments of orthologous intergenic regions for short segments that are significantly more conserved than expected based on the phylogeny of the species.
In this work the authors present an algorithm, PhyloGibbs, that combines these two approaches into one integrated Bayesian framework. The algorithm searches over all ways in which an arbitrary number of binding sites for an arbitrary number of transcription factors can be assigned to arbitrary collections of multiple sequence alignments while taking into account the phylogenetic relations between the sequences.
The authors perform a number of tests on synthetic data and real data from Saccharomyces genomes in which PhyloGibbs significantly outperforms other existing methods. Finally, a novel anneal-and-track strategy allows PhyloGibbs to make accurate estimates of the reliability of its predictions.
Citation:Siddharthan R, Siggia ED, van Nimwegen E (2005) PhyloGibbs: A Gibbs sampling motif finder that incorporates phylogeny. PLoS Comput Biol 1(7): e67.
==== Body
Introduction
Transcription factors (TFs) are proteins that bind in a sequence-specific manner to short DNA segments (“binding sites”), most commonly in intergenic DNA upstream of a gene, to activate or suppress gene transcription. Their DNA-binding domains recognize collections of short related DNA sequences (“motifs”). One generally finds that, although there is no unique combination of bases that is shared by all binding sites, and although different bases can occur at each position, there are clear biases in the distribution of bases that occur at each position of the binding sites. A common mathematical representation of a motif that takes this variability into account is a so-called weight matrix (WM) [1,2] w, whose components w
αi give the probabilities of finding base α ∈ {A, C, G, T} at position i of a binding site. The main assumption underlying this mathematical representation is that the bases occurring at different positions of the binding site are probabilistically independent. This in turn follows, under some conditions [3], from the assumption that the binding energy of the protein to the DNA is a sum of pairwise contact energies between the individual nucleotides and the protein.
There are several algorithms that are based on the WM representation that detect, ab initio, binding sites for a common TF in a collection of DNA sequences [4–7]. These algorithms broadly fall into two classes. One class, of which MEME [6] is the typical representative, searches the space of all WMs for the WM that can best explain the observed sequences. The class of “Gibbs sampling” algorithms, of which the Gibbs motif sampler [4,5] is the typical representative, instead samples the space of all multiple alignments of small sequence segments in search of the one that is most likely to consist of samples from a common WM.
A crucial factor for the success of ab initio methods is the ratio of the number of binding sites to the total amount of DNA in the collection of sequences. That is, the larger the number of binding sites in the set, and the smaller the total amount of DNA, the more likely it is that ab initio methods can discover the binding sites among the other DNA sequences. In order to ensure a reasonable chance of success one thus needs to provide these methods with collections of sequences that are highly enriched with binding sites for a common TF. One possibility is to use sets of upstream regions from genes that appear co-regulated in microarray experiments (e.g., [8,9]) or that were bound by a common TF in ChIP-on-chip experiments (e.g., [10]). Another possibility is to use upstream regions of orthologous genes from related organisms. Here the assumption is that (most of) the regulation of the ancestor gene, and thus its binding sites, has been conserved in the orthologs that descend from it.
This latter approach is in general complicated by a number of factors. When searching for regulatory sites in sequences that are not phylogenetically related, such as upstream regions of different genes from the same organism, one may simply look for short sequence motifs that are overrepresented among the input sequences. If the set of species from which the orthologous sequences derive are sufficiently diverged, one may simply choose to ignore the phylogenetic relationship between the sequences and treat the orthologous sequences in the same way as sequences that are not phylogenetically related. This was, for instance, the approach taken by McCue et al. [11,12], where the Gibbs motif sampler algorithm [4,5] was used on upstream regions of proteo-γ bacteria.
However, this approach is not applicable to datasets containing more closely related species, where some of the sequences will exhibit significant amounts of similarity simply because of their evolutionary proximity. Moreover, the amount of similarity will depend on the phylogenetic distance between the species, and it is clear that finding conserved sequence motifs between orthologous sequences from closely related species is much less indicative of function than finding sequence motifs that are conserved between distant species. One will in general thus have to distinguish conservation due to functional constraints from conservation due to evolutionary proximity, and to do this correctly, the phylogenetic relationship between the sequences has to be taken into account.
A second challenge in using orthologous intergenic sequences from multiple species is the nontrivial structure of their multiple alignments. One typically finds a very heterogeneous pattern of conservation: well-conserved blocks of different sizes and covering different subsets of the species are interspersed with sequence segments that show little similarity with the sequences of the other species.
The technique of “phylogenetic footprinting” (e.g., [13–17]), restricts attention to only those sequence segments in the genome of interest that show significant conservation with the other species. The conserved regions for multiple genes are then searched for common motifs by a variety of techniques. It is unclear, however, to what extent regulatory sites are restricted to such conserved segments. For instance, several studies of Drosophila and yeast [18–20] have shown that there is no strong correlation between where experimentally annotated binding sites occur and whether that region is conserved. Thus, at least for yeast and flies, considerable information is lost by focusing on the conserved regions only.
We thus decided to retain the entire patchwork pattern of conserved sequence blocks and unaligned segments. Our strategy is implemented by a Gibbs sampling approach, and a preliminary account of the algorithm was presented in [21]. The algorithm operates on arbitrary collections of both phylogenetically related sequences, such as orthologous intergenic regions, and sequences that are not phylogenetically related, such as upstream regions of different genes from the same organism. The phylogenetically related groups of sequences in the input are pre-aligned into local multiple alignments where clearly similar sequence segments are aligned into blocks and sequence segments of no or marginal similarity are left unaligned [22]. Although the algorithm can also take global multiple alignments as input, we believe that these often force phylogenetically unrelated segments into aligned blocks. This may adversely affect the performance of the algorithm. We score putative sites within blocks of aligned sequences with an evolutionary model that takes the phylogenetic relationships of the species into account, while putative sites in unaligned segments are treated as independent occurrences. This Bayesian model defines a probability distribution over arbitrary placements of putative binding sites for multiple motifs, and we sample it with a Monte Carlo Markov chain. We first use simulated annealing to search for the globally optimal configuration of binding sites. The motifs in this configuration (which hopefully corresponds to the global optimum) are then “tracked” in a further sampling run to estimate realistic posterior probabilities for all the binding sites that the algorithm reports.
Recently a number of other algorithms have been developed that search for regulatory motifs in groups of phylogenetically related sequences. Probably the first algorithm that was proposed is a generalization of the Consensus algorithm [7] called PhyloCon [23]. PhyloCon operates on sets of co-regulated genes and their orthologs. It is a greedy algorithm that first finds ungapped alignments of similar sequence segments in sets of orthologous sequences, and then combines these alignments from different upstream regions into larger alignments. This algorithm does not take any phylogenetic information into account, i.e., closely related sequences are treated the same as distantly related sequences. Other drawbacks of this algorithm are that it assumes that each motif will have exactly one site in each of the intergenic regions and that it assumes that this site is conserved in all orthologs.
More closely related to PhyloGibbs's approach are two recent algorithms [24,25] that generalize MEME [6] to take the phylogenetic relationships between species into account. The main difference between EMnEM and PhyME is that PhyME uses the same evolutionary model for the evolution of binding sites as PhyloGibbs, which takes into account that binding sites evolve under constraints set by a WM, whereas EMnEM simply assumes an overall slower rate of evolution in binding sites than in background sequences. Another difference is that PhyME, like PhyloGibbs, treats the multiple alignment more flexibly than EMnEM, which demands a global multiple alignment. The main difference between PhyloGibbs and these algorithms is of course that PhyloGibbs takes a motif sampling approach, which allows us to search for multiple motifs in parallel, whereas PhyME and EMnEM use expectation maximization (EM) to search for one WM at a time.
In the following sections, we first describe our Bayesian model that assigns a posterior probability to each configuration of binding sites for multiple motifs assigned to the input sequences. We start by describing the model for phylogenetically unrelated sequences, which is essentially equivalent to the model used in the Gibbs motif sampler [4,5], and then describe how this model is extended to datasets that contain phylogenetically related sequences. After that we describe the move set with which we search the state space of all possible configurations, and the annealing and “tracking” strategy that we use to identify the significant groups of sites.
We then present examples of the performance of ours and other algorithms on both synthetic and real data. The synthetic datasets consist of mixtures of WM samples and random sequences, which is in accordance with assumptions that all algorithms make. This allows us to compare the performance of the algorithms in an idealized situation that does not contain the complexities of real data. These tests also show to what extent binding sites can be recovered for this idealized data as a function of the quality of the WMs, the number of sites available, and the number of species available and their phylogenetic distances. For our tests on real data we use 200 upstream regions from Saccharomyces cerevisiae that have known binding sites from the collection [26], and compare the ability of the different algorithms to recover these sites when running on multiple alignments of the orthologs of these upstream regions from recently sequenced Saccharomyces genomes [15,16]. Finally, we run PhyloGibbs on collections of upstream region alignments that were annotated in [27] to contain binding sites for a common TF based on data from ChIP-on-chip experiments, and we extensively compare PhyloGibbs' results with the annotations in [27] and with the literature.
Results
Model for Phylogenetically Unrelated Sequences
In order to motivate and explain our model for phylogenetically related sequences it is helpful to first introduce the model for sequences that are not phylogenetically related. In this context, “not phylogenetically related” means that for any pair of sequences in the input data, their common ancestor sequence is sufficiently far in the evolutionary past that mutations have been introduced multiple times at each position in the sequences. That is, any similarity left between the input sequences cannot be due to evolutionary proximity.
We assume that our data contain an unknown number of sites for an unknown number of different TFs. The state space of possible solutions to the problem of identifying the binding sites contained in these sequences consists of all possible ways in which one can assign groups of binding sites to these sequences. An example of such binding site assignments, which we call “configurations,” is shown in Figure 1.
Figure 1 Binding Site Configuration
A window, in our terminology, is a possible binding site for a TF; in the case of phylogenetically unrelated sequences it is simply a set of m contiguous bases in a sequence, with m the binding site width. This figure shows a configuration C containing a total of eight windows (rectangles) for three different WMs (red, blue, and green). Note that a single sequence of length L has L − m + 1 windows in it.
Assuming that the width of the binding sites is m bases, each contiguous segment of m bases in our dataset is a potential binding site. We call such m-base segments “windows” and will generalize this concept later for phylogenetically related sequences. We label the WMs for the different TFs with “colors” one, two, etc., and use the color zero to indicate nonfunctional “background” sequence. Thus, formally, a configuration C is an assignment of nonzero colors to a particular set of nonoverlapping windows. Note that the requirement that colored windows cannot overlap means that a colored window “blocks” up to 2(m −1) others.
Given a dataset of input sequences S, our model assigns to each possible configuration C a posterior probability P(C|S). Using Bayes's theorem we have
where P(C) is the prior probability of configuration C. As described in Materials and Methods, PhyloGibbs provides for priors that incorporate prior information on the likely number of motifs and binding sites in the input sequences. The likelihood function P(S|C) gives the probability that, for each color in C, all sequences of the same color were sampled from a common WM, multiplied by the probability of the remaining sequences under a “background” model. That is, we formally have
where we denote all sequence that is colored zero in the configuration C by S ∉ C, the background model for this sequence by P(S ∉ C|B), the set of sequences in windows with color c as Sc, and the probability that all sequences in Sc were drawn from a common WM by P(Sc). Expressions for all these quantities are derived in Materials and Methods. The probability P(Sc) that all sequences in color c were sampled from the same WM is obtained by taking the probability P(Sc|w) that all sequences Sc were sampled from a particular WM w, and integrating this over all possible WMs w [28].
Incorporating Phylogeny
When considering datasets that contain phylogenetically related sequences, such as orthologous intergenic regions from related species, the main problem is distinguishing sequence similarity that is due to evolutionary proximity from sequence similarity that arises from functional constraints. That is, when calculating the probability P(S) that a group of sequences S are all binding sites for a common WM, we should treat sequences that are orthologous by a different model than those that are not phylogenetically related. We derive such a model below. In order to apply this model we also have to determine which parts of the input sequences are orthologous, i.e., we need to provide a multiple alignment. We could in principle let the algorithm search both the space of multiple alignments and the space of binding site configurations C at the same time, but we decided that this state space is likely too large to search effectively. Moreover, for closely related species large segments of the orthologous intergenic regions can be unambiguously aligned, and by pre-aligning these we significantly reduce the search space for the algorithm.
Our strategy is thus to first produce a multiple alignment and then search the space of binding site configurations that are consistent with this alignment. Standard global multiple alignment algorithms [29–31] can be used for this task, and PhyloGibbs can be run on the outputs they produce. However, as discussed in the Introduction, alignments of orthologous intergenic regions from related species (such as the recently sequenced Saccharomyces species [15,16]) show a mosaic pattern of well-conserved blocks interspersed with stretches of unconserved sequence, and global alignment algorithms may spuriously align many of these phylogenetically unrelated parts. Binding sites are typically not restricted to conserved regions [18,19], and spurious alignment of phylogenetically unrelated regions may hamper the discovery of binding sites contained within them. Our strategy is thus to only align those parts of the intergenic regions that are clearly orthologous and can be unambiguously aligned, and to leave the rest unaligned. In searching binding site configurations we demand consistency of the configuration with the aligned orthologous blocks, but at the same time also allow windows to be placed in unaligned parts of the sequences. Such syntenic local multiple alignments are produced by the algorithm Dialign [22], and we used this algorithm for producing the multiple alignments of the datasets that we report on below (see [32] for a recent review of the performance of various alignment algorithms on orthologous intergenic DNA). Another algorithm that produces syntenic local multiple alignments and is especially suited for large genomic regions is Threaded Blockset Aligner [33].
Once we have a syntenic multiple local alignment, we treat columns of aligned bases as phylogenetically related, i.e., arising from a common ancestor base. The state space again consists of all possible configurations of binding sites but now with the constraint that “windows” that include aligned bases have to extend over all sequences in the alignment. That is, we assume that if a binding site occurs in a sequence segment that is aligned with sequence segments from the other species, then binding sites for the same TF have to occur in the corresponding positions of these other sequence segments. To this end we extend the concept “window” (denoting a position of a potential binding site) to multiple local alignments, as illustrated in Figure 2.
Figure 2 An Alignment of Four Sequences Showing Three Legitimate Windows and One Illegitimate Window
Vertically aligned capital letters are phylogenetically related bases, assumed to have evolved from a common ancestor. Thus, any window placed on these bases is extended to cover all related bases. Three legitimate windows are surrounded by solid boxes. The window surrounded by the dotted box is illegitimate because the gap in the top sequence makes the alignment of bases inconsistent. Note that lower case letters are not aligned and that, in order to complete a window with aligned sequences, one may slide lowercase bases “through” adjoining gaps. For example, if the window on the bottom two sequences were to move two steps to the left, the “c” and “a” on the left side of the preceding gaps would slide through the gaps to the right to complete the window.
The figure shows a sample stretch of four aligned sequences, where uppercase letters are aligned and lowercase letters are “independent.” In an initial pass the program identifies the set of all legitimate “windows” in the entire sequence data. Each of these windows may encompass one or more sequences. The windows must contain consistently aligned uppercase letters: there should not be “gaps” that give inconsistent spacing between aligned uppercase letters. For example, in Figure 2, the sequences boxed with solid lines show consistent windows, whereas a dotted line boxes an inconsistent window. Note that, as in the leftmost window in the figure, lowercase letters can be used to complete a window in which only some letters are uppercase. The details of identifying the set of all legitimate windows in the sequence data are described in Materials and Methods. At the end of this procedure, we have a list of windows that represent potential sites for TF binding sites; some of these windows contain only one sequence and represent a potential independent occurrence of a binding site, while others contain multiple sequences and represent potential binding sites that evolved from a common ancestor site and were conserved across multiple species.
Next, we need to generalize our probabilistic model to multiply aligned orthologous sequences. For the single-sequence windows of the previous section, the probability P(s|w) of observing the sequence s given a WM w is simply given by
with si the base at position i of s. For a window overlaying a region of multiply aligned sequences, the single base si is replaced by the set of bases Wi in the ith column of the alignment. We now replace the probability
with the probability P(Wi|w) that the bases Wi in the ith column of the window derive from a common ancestor base, and have been evolving under the selective constraints set by the requirement that they remain binding sites for the TF represented by WM w. Our evolutionary model assumes that mutations are introduced at a fixed rate and that the probability for selection to fix a mutation toward base α in the ith column of a site is proportional to the WM entry w
αi. Since the positions are mutually independent we have for the whole window
The probability P(Sc|w) that all windows W ∈ c in color c evolved under the constraints set by WM w is simply given by the product
and the probability P(Sc) is again given by an integral of P(Sc|w) over the space of all possible WMs. The background score P(W|B) for a window W is given by the exact same expression as P(W|w) except that the WM entries w
αi are replaced with the background probabilities for the bases in each column. Detailed derivations and explicit expressions are provided in Materials and Methods.
Move Set
The last two sections have explained the posterior probability P(C|S) that we assign to configurations of binding sites C given the input data S. PhyloGibbs samples the space of all possible configurations C using a Monte Carlo Markov-chain sampling strategy [34] using a move set with a number of different moves. Generally, the moves in our move set operate on configurations either by shifting one or more colored windows in the configuration, or by adding/removing a colored window to/from the configuration. Formally, a “move” consists of taking the current configuration C, constructing the set X of all configurations C′ ∈ X that differ from C by a “single change” (e.g., moving the position of a single colored window), and then choosing one of these configurations C′ according to the distribution
. In order for the repeated application of these moves to sample the whole space of configurations in proportion to their probabilities P(C|S), two conditions are sufficient. First, the move set needs to be “ergodic”: each configuration C must be reachable by repeated application of the moves in the move set. And second, we demand that “detailed balance” be satisfied: for any pair of states C and C′ the probabilities P(C → C′) of moving from C to C′ and P(C′ → C) of moving from C′ to C under any of the moves must satisfy
The most important of the moves in our move set is the “window shift” move, which takes a single window and resamples its position. Since this type of move is generally referred to as Gibbs sampling, i.e., one samples a joint probability distribution by resampling one variable of the joint distribution at each time step while keeping the other variables fixed, and because of the similarities with the original Gibbs motif sampling algorithm [4,5], we have called our algorithm PhyloGibbs. Other moves in our move set include picking a window at random and recoloring it, and shifting all windows in a color by the same amount. Each time step of the algorithm corresponds to a “cycle” of a fixed number of moves of each type. The moves in our move set were designed to avoid the algorithm getting trapped in local maxima, to speed up convergence, and to ensure ergodicity. They are described in detail in Materials and Methods.
Strategy: Anneal and Track
By repeating moves from the move set described in the previous section PhyloGibbs will, in the limit of long time, sample each configuration C according to its posterior probability P(C|S). Even though the distribution P(C|S) represents everything that can be inferred from the data using our model, we still have to decide how to usefully summarize this information. There is, to our knowledge, currently no generally satisfactory solution to this problem (see the discussion in [28]). One would like the summary to report all the relevant features that are shared by the configurations C with high posterior probability P(C|S). In particular, one would like to identify groups of windows that with high probability share a color. Given a reference set of windows, it is straightforward to check what fraction of the time different subsets from this reference set are colored the same, and the fraction of the time that other windows are co-colored with windows in this reference set. However, one obviously cannot check this for all possible subsets of windows, and we thus have to decide what reference sets we want to “track.”
Our current strategy is to first search for the configuration C* that has the globally maximal posterior probability P(C|S) and to take the sets of co-colored windows in this configuration as the reference sets. We use simulated annealing to search for this globally optimal configuration C*, which we call the “reference configuration.” Instead of sampling configurations from the distribution P(C|S) we raise each probability to the power β and sample from the posterior P̃(C|S) ∝ P(C|S)β. Initially we set β = 1 and slowly increase β with time until the sampler “freezes” into a configuration C* with (at least locally) maximal probability P(C*|S).
The annealing phase is followed by a tracking phase in which we sample from the distribution P(C|S) in order to estimate, for each window w and each color c, the probability p(w,c) that window w belongs to the same motif (color) as the windows in color c of the reference configuration C*. The probabilities p(w,c) are estimated as follows. After every cycle of moves we compare the current configuration C with the reference configuration C*. Specifically, for every color c in the reference configuration C* we determine which color c̃ of the current configuration C matches c most closely (allowing for shift and reverse-complement of the windows in c̃ with respect to the windows in c; see Materials and Methods). For every window w in c̃ (properly shifted to align with c), we add a count of one to the number of times n(w,c) that w is associated with color c. At the end we set p(w,c) = n(w,c)/n, with n the total number of cycles.
In its default mode of operation PhyloGibbs reports the following results for the input set of sequences S: (1) the configuration C* that has maximal posterior probability P(C|S), (2) an inferred WM for each color c in configuration C*, (3) for each color c, all windows w for which p(w,c) ≥ p
min, with p
min a cutoff that the user can specify, and (4) a WM for each color c from reference configuration C* that is inferred by weighing the member windows w with their membership probabilities p(w,c).
Note that, in general, different member windows w of color c in reference configuration C* will have very different posterior probabilities p(w,c) to be members of the motif. In addition, for each color c the tracking will typically uncover windows w that were not a member of color c in configuration C* but that with reasonable probability p(w,c) belong to the motif as well. One inherent limitation of our anneal-and-track strategy is that it can in principle miss a group of windows that are co-colored a significant fraction of the time in P(C|S) but that do not occur in configuration C*. Trivially, if the user allows for too few colors, some motifs may be missed. It generally does not hurt the results to overestimate the number of colors (different motifs), although it increases running time. Similarly, it is also not necessary to accurately estimate the total number of sites. If the prior overestimates the total number of sites, some spurious sites in the reference configuration will simply fail to track. Similarly, if the total number of sites is underestimated, some sites that were missed in the reference configuration will be picked up in the tracking phase. To give maximal flexibility to the user, the algorithm can also, instead of doing an anneal, be given an input reference configuration and track the motifs in this configuration.
As far as we are aware, PhyloGibbs is the only motif-finding algorithm that rigorously assigns posterior probabilities p(w,c) to the binding sites that it reports by sampling the entire space of binding site configurations. The Gibbs motif sampler [4] WGibbs samples the space of configurations while keeping track of the best configuration C* that it has seen, i.e., it does not use annealing. It also reports posterior probabilities for the sites it reports, but these are based on sampling of only a small subspace of configurations around the configuration C*. Algorithms that search the space of WMs [6,25,24] use EM to find an optimal WM explaining the data and then assign posterior probabilities to reported sites by assuming this WM is correct. That is, they do not take into account the (likely) possibility that the WM found through EM is not entirely correct.
As we show below, only the rigorous sampling of the space of all configurations as implemented in PhyloGibbs is capable of assigning realistic posterior probabilities to the sites it reports. It is sometimes attempted to identify “significant” motifs by simply rerunning one or several different motif-finding algorithms and looking for recurring motifs. However, this merely generates the subsidiary problem of clustering the multiple predictions. Instead of using ad hoc scoring schemes for clustering, reported binding sites should ideally be clustered using the same probabilistic scoring that generated them, i.e., as in [28]. PhyloGibbs generalizes the binding site clustering algorithm of [28] and subsumes any problems of clustering.
Performance on Synthetic Data: Anneal
In general there are three qualitatively different issues that contribute to the performance of motif-finding algorithms on real data. First, all motif-finding algorithms make assumptions about the data that will ignore at least some of the complexities of real data. The performance of a given motif-finding algorithm will depend on the extent to which these ignored complexities affect the algorithm's ability to perform its task. Second, the search spaces of all possible WMs or all possible binding site configurations are too large to search exhaustively, and therefore all algorithms employ heuristic methods to search for the globally optimal WMs or configurations. The extent to which the heuristic methods succeed or fail will also affect the performance of the algorithms. Third, even if the data adhere to all assumptions that an algorithm makes, and the algorithm successfully finds the global optimum in the search space, this still does not guarantee that the algorithm will recover the correct motifs and sites. That is, if the motifs are fuzzy and the sites are embedded in long background sequences it might occur that, by chance, the background contains sets of sites that are more conserved and more similar than the embedded sites. In this case it will be impossible for any algorithm to recover the true sites.
By generating synthetic data to accord, as much as possible, with the assumptions that the motif-finding algorithms make, we can study the second and third issues separately from the first issue. In this section and the next we do a number of such tests. In our first test we want to evaluate to what extent PhyloGibbs can recover a fixed number of sites embedded in a perfect alignment of orthologous sequences as the quality of the WMs and the phylogenetic distances of the orthologs are varied. At the same time, we want to test how well PhyloGibbs performs when operating on perfect alignments compared to algorithms that do not take phylogenetic information into account and that cannot operate on multiple alignments (including PhyloGibbs in the mode where it ignores phylogenetic information). This test will indicate how much performance can be improved by using phylogenetic information and multiple alignments in an ideal situation. For ease of reference, from now on we refer to all algorithms that use phylogenetic information and that can operate on multiple alignments as “phylo” algorithms, while referring to algorithms that treat all sequences as independent as “non-phylo” algorithms.
For our first test we generated synthetic datasets as follows. (1) We first generated a WM of width w. For each position in the WM we picked a random “consensus” base, set the probability of that base to p, and set the probabilities of the other bases to (1 − p)/3. The WM was thus parametrized by p, which we call its “polarization.” For some datasets we also generated random WMs by picking, for each position, a random distribution (wa, wc, wg, wt) uniformly from the simplex
. Note that for these random WMs the “polarization” varies from column to column. (2) We sampled s sites from the WM and embedded them in a random sequence of length L. This sequence formed the ancestor sequence. (3) We then generated S descendant sequences as follows. For each base the descendant's base was equal to the ancestor's base with probability q. With probability 1 − q we mutated the base. Outside binding sites, a mutation replaced the ancestor base with a randomly chosen base. Within binding sites, a mutation replaced the ancestor base with a new sample from the WM. Because the WM is generally biased toward particular bases this results in more conservation within binding sites than outside them. We refer to the parameter q as the “proximity” of the descendants to their common ancestor. (4) We measured performance by the overlap between the predicted sites in the reference configuration C* and the embedded sites. Formally, we counted the number of bases in the intersection of predicted and embedded sites and divided by the total number of bases in embedded sites (which, in these tests, equals the total number of bases in predicted sites). A performance of one thus corresponds to a perfect overlap of the embedded and predicted sites.
We compared the performance of PhyloGibbs with those of non-phylo algorithms on alignments of S = 5 orthologous intergenic regions of length L = 500 containing s = 4 sites for a single motif, as a function of the WM polarization p and the proximity of the descendants q. The results are shown in Figure 3.
Figure 3 Performance of PhyloGibbs and Non-Phylo Motif-Finding Algorithms on Alignments of Orthologous Intergenic Regions as a Function of the Evolutionary Proximity of the Orthologs and the Quality of the WM
PhyloGibbs with phylogeny (red), PhyloGibbs in non-phylo mode (light blue), WGibbs (dark blue), and MEME (pink) were run on alignments of S = 5 intergenic regions of length L = 500, each at a proximity q to the common ancestor and each containing s = 4 binding sites from a single WM of width w = 10. In the upper left panel, WMs had polarization p = 0.6, in the upper right p = 0.75, in the lower left p = 0.9, and in the lower right random WMs (drawn uniformly from the simplex) were used. The solid lines show the average overlaps between the predicted sites and the real sites, and the dotted lines show two standard errors (estimated from 50 different datasets generated with equal parameters for each data point).
All algorithms assume that the data are a mixture of random uncorrelated background sequences and samples from a number of WMs of certain lengths. With the exception of the phylogenetic relationship of the sequences, which is ignored by the non-phylo algorithms, the synthetic data are thus in complete accordance with the assumptions that each of the algorithms make. For each algorithm we specified the correct length and number of sites. Since when using PhyloGibbs with phylogeny the windows extend over all five sequences in the alignment, we asked PhyloGibbs to predict four multi-sequence windows for a single motif, while we asked the non-phylo algorithms to search for 20 single-sequence sites for a single motif. Since for any algorithm, the performance differs substantially between input datasets that were generated with the same parameter settings, we averaged results over 50 datasets and in Figure 3 show two standard errors (dotted lines) around the average performance (solid lines).
All non-phylo algorithms, including PhyloGibbs when phylogeny is turned off, perform roughly equally well (or badly). For highly polarized WMs all non-phylo algorithms perform quite well. In contrast, for low polarizations (p = 0.6 in the upper left panel) or for random WMs (lower right panel), all non-phylo algorithms perform hardly better than random predictions would do (the four embedded binding sites cover 8% of the input data). The second thing to note is that, especially as phylogenetic distance between the orthologs increases (lower q), PhyloGibbs performs substantially better than the non-phylo algorithms. As the amount of conservation due to evolutionary proximity becomes larger than the amount of conservation due to functional constraints, i.e., as q becomes larger than p, the performance drops and the performance of PhyloGibbs becomes only marginally better than that of the non-phylo algorithms.
It is important to point out that PhyloGibbs's superior performance for these data is partly due to the fact that the five sequences have been perfectly aligned and that it is searching only through configurations that are consistent with this alignment. In contrast, the non-phylo algorithms treat the five sequences as independent and have to search a much larger space of configurations. For real data we of course do not have perfect alignments and it will generally be hard to obtain good alignments when the proximity q becomes small, which will negatively affect the performance of PhyloGibbs. This effect is seen below in our tests on real data.
In Figure S1 we show the results of additional tests analogous to those shown in Figure 3. In these tests we chose the WMs and phylogeny to mimic as closely as possible the situation in yeast, whose real data we study below. That is, instead of creating random WMs we used “real” WMs of yeast TFs with known binding specificities, and we used phylogenetic distances for the five descendants that are proportional to the phylogenetic distances of the Saccharomyces sensu stricto species that we use in our tests on real data below. The results with these more realistic synthetic datasets are quantitatively very similar to the results shown in the upper right panel of Figure 3.
It might also be asked if the non-phylo algorithms are put at a disadvantage by the fact that they have to search for a much larger number of sites. That is, to get a 50% performance PhyloGibbs needs only to get two multi-species sites correct, whereas the non-phylo algorithms need to get ten sites correct. To test this we ran MEME and WGibbs on single sequences, as opposed to groups of S = 5 orthologs, and asked the algorithms to find only s = 4 sites instead of sS = 20. The non-phylo algorithms MEME and WGibbs performed extremely poorly in this mode. It thus appears that, at least for this kind of synthetic data, the benefit of having more sites outweighs the negative effects of having more sequence to search through.
Although PhyloGibbs performed consistently better than the non-phylo algorithms, in many cases it recovered only a fraction of the embedded sites. Since the synthetic data were generated exactly according to the model that PhyloGibbs assumes, there are only two possible reasons for the failure of PhyloGibbs to recover the embedded sites. The first possibility is that the correct configuration, i.e., with the four binding sites occurring at the positions where they were embedded, is the globally optimal binding site configuration, but that the anneal phase failed to identify it and instead settled on an only locally optimal configuration. In that case the posterior probability P(C
cor|S) of the correct configuration C
cor should be higher than the posterior probability P(C*|S) of the configuration C* that the anneal obtained. The second possibility is that the anneal did identify the globally optimal configuration, and that this configuration C* has higher posterior probability P(C*|S) than the posterior probability of the correct configuration P(C
cor|S). This can happen when, by chance, positions outside of the embedded binding sites are more conserved and/or better match a common WM than the embedded sites, making it impossible for any algorithm to recover them correctly.
To investigate how often the anneal in PhyloGibbs identifies the globally optimal configuration, we compared P(C*|S) with P(C
cor|S) for the runs with proximities q = 0.2, q = 0.5, and q = 0.8, shown in the lower right panel of Figure 3. For each value of q there were 50 independent datasets generated, and PhyloGibbs was run five times on each dataset. Thus, for each value of q there were 250 runs in total. We found that P(C*|S) ≥ P(C
cor|S) for 245 of the 250 runs for q = 0.2, for 243 of the 250 for q = 0.5, and for all 250 runs for q = 0.8. Thus, for 98.4% of the runs, the posterior probability of the configuration found in the anneal was at least as high as the posterior probability of the correct configuration.
In conclusion, these first tests with synthetic data showed that, when sites from WMs are embedded in a random ancestor sequence, and PhyloGibbs is given a perfect alignment of a set of descendants of this sequence, it performs significantly better than algorithms that treat the descendants as independent sequences. It also shows that as the similarity between sites becomes less than or equal to the similarity between orthologous sequences due to evolutionary proximity, it becomes impossible for any algorithm to accurately recover the sites.
In the first test PhyloGibbs used both information from the overrepresentation of a motif in the data and information about the conservation of its sites. We next investigated how many species one would need, in an ideal situation, to reliably infer the location of a single binding site using conservation only. To test this we generated synthetic intergenic regions of length L = 500 that contained a single site for a single randomly chosen WM of width w = 10 and created alignments of S descendant sequences with proximity q = 0.5. We then ran PhyloGibbs for different values of S, asking it to search for a single multi-sequence window in the alignment. The results are shown in Figure 4.
Figure 4 Performance of PhyloGibbs in Recovering a Single Site of a Randomly Chosen WM of Width w = 10 from the Alignment of S Orthologous Intergenic Regions of Proximity q = 0.5 and Length L = 500 as a function of S
The solid line shows the average overlap between the true site and the predicted site and the dotted lines show two standard errors.
We see that more than ten species are needed to have a 50% probability to recover a single site of a random WM at q = 0.5, and that as many as 25 species are necessary to recover the site with 90% probability. Note that significantly more species are necessary to recover a single site than are necessary to recover a group of multiple sites from the same WM. That is, in the lower right of Figure 3 about 40% of the quartet of sites is recovered at q = 0.5, whereas in Figure 4 at S = 5 the single site is recovered with a probability of only 0.15. In conclusion, this test has shown that if enough species are available in which a given binding site occurs, and these can be reliably aligned, then even individual sites can be reliably recovered by PhyloGibbs.
Performance on Synthetic Data: Tracking
In the next section we compare the performance of PhyloGibbs and other algorithms that use phylogeny (PhyME [25] and EMnEM [24]) on real data from five Saccharomyces species. To compare and contrast with those tests we here create synthetic datasets that are comparable with those real datasets but that are idealized in the sense that they respect the assumptions that the various algorithms make about the data as much as possible. This test allows us to directly compare the performance of the phylo algorithms on idealized data, and the extent to which they outperform non-phylo algorithms in this idealized setting.
As mentioned in the previous section, our synthetic data are in accordance with all assumptions that the non-phylo algorithms make about the data, except of course for the phylogenetic relationships between the sequences. Our synthetic orthologous sequences are generated in accordance with the evolutionary model that PhyloGibbs and PhyME assume. For these two algorithms the synthetic data are thus in exact accordance with the assumptions that these algorithms make. EMnEM employs an evolutionary model that uses the same substitution matrix both within and outside of sites, but allows each position in a binding site to evolve at a different overall rate. This model is thus less realistic than the model that PhyME and PhyloGibbs use in that it ignores that the probabilities of different substitutions within a site depend on the site's WM. However, since it has more free parameters that can be fitted, we suspect that in practice it will be able to reasonably approximate the evolutionary model that PhyloGibbs and PhyME use.
We followed the estimates of [35] and created 250 synthetic datasets, each containing nine binding sites, sampled equally from three random WMs for each intergenic region (see Figure 5 for details). We assessed the results with a plot of specificity (fraction of predictions matching true sites) versus sensitivity (fraction of true sites that were recovered). All the algorithms report posterior probabilities (or a p-value) for the sites they report, which we used to rank the predictions. We pooled the predictions from all datasets and then generated successively larger lists of predictions by including all predictions over a given posterior probability. For each list we then determined the specificity and sensitivity and plotted them in Figure 5 (see Materials and Methods for details).
Figure 5 Performance of Several Motif-Finding Algorithms on Synthetic Data Prepared as for Figure 3
A total of 250 alignments of S = 5 orthologous intergenic regions of length L = 750 and proximity q = 0.5 were created with three binding sites sampled from each of three different random WMs. The left panel shows how the fraction of predicted sites that match true sites (specificity) depends on the fraction of true sites that are among the predictions (sensitivity) for PhyloGibbs (red), EMnEM (yellow), PhyME (green), PhyloGibbs without phylogeny (light blue), WGibbs (dark blue), and MEME (pink). Dashed lines correspond to two standard errors. The right panel shows the ability of the different algorithms to assess their own reliability. The true specificity is shown as a function of the specificity that the algorithm predicts for the sites that it reports. The black line y = x corresponds to a perfect assessment of the algorithm's reliability.
The algorithms that ignore phylogeny did not recover more than 16% of the true sites (sensitivity), and did so with a nearly fixed specificity of around 20%, meaning that there is not much enrichment in true sites in the top versus the bottom of their ranked lists. The algorithms that exploit the phylogeny all did better for the simple reason that they operate on the perfect multiple alignments and therefore their search space is much smaller. Of these three algorithms PhyloGibbs performed best. PhyME, in common with the non-phylo algorithms, reports a very limited range of posterior probabilities for the sites it reports, which leads to a relatively small “dynamic range” in sensitivity/specificity.
Note that even with this perfectly aligned data, to get 50% of the true sites, PhyloGibbs needed to make more than twice as many predictions. Again, to determine to what extent the failure of PhyloGibbs to recover all the embedded sites was caused by the anneal getting trapped in locally optimal configurations, we compared the posterior probabilities P(C*|S) of the reference configurations with those of the correct states P(C
cor|S). We found that P(C*|S) was greater than or equal to P(C
cor|S) for all 250 datasets.
These synthetic data also provide the opportunity to test how well the algorithms assess their reliability, i.e., how well the reported posterior probabilities for their predictions match the specificities (fraction of predictions matching true sites) we compute by knowing the true sites. Ideally the two are the same, so that for real data one could use the posterior probabilities to gauge the fraction of correct predictions. The right panel of Figure 5 shows that with the exception of PhyloGibbs (with and without phylogeny) all algorithms were much too optimistic about the quality of their predictions: EMnEM and MEME gave posterior probabilities larger than 95% when their specificity was around 35%, and WGibbs gave posteriors of 90% when its real specificity was only 20%. MEME is not included in the right panel of Figure 5 because it reports p-values instead of posterior probabilities.
Both EMnEM and PhyME overestimated their specificity because they calculate their posterior probabilities for sites under the assumption that the WMs that they infer are correct. In reality, the inferred WM will often not match the true WM that generated the data. For WGibbs the overestimation stems from the restricted sampling of configurations around the one that gave the maximum posterior probability during sampling. Only PhyloGibbs bases its posterior on sampling of the whole space of binding site configurations.
In Figure S2 we present the results of a test analogous to the one in Figure 5, using again “real” WMs from yeast instead of random WMs, and using the proximities of the sensu stricto Saccharomyces species instead of proximity 0.5 for all descendants. All algorithms performed substantially better in these tests, which is in all likelihood mostly because of the higher information scores (see Materials and Methods) of the yeast WMs compared to random WMs. All phylo algorithms still outperformed the non-phylo algorithms, and of the phylo algorithms, PhyloGibbs performed significantly better than PhyME and EMnEM. In contrast to the test shown in Figure 5, PhyME clearly outperformed EMnEM on this test.
In summary, these tests have shown that, given perfectly aligned input sequences, all phylo algorithms substantially outperform non-phylo algorithms. The tests have also shown that, on data that are in accordance with the assumptions that the phylo algorithms make (almost all for EMnEM), PhyloGibbs outperforms the other algorithms. In addition, only PhyloGibbs is capable of reasonably estimating the reliability of its own predictions.
Results on the Yeast Genome
To test the performance of PhyloGibbs and other algorithms on real data we decided to use data from the recently sequenced yeast genomes [15,16]. For our first test on these data we used the list of documented binding sites for yeast TFs in the Saccharomyces cerevisiae promoter database (SCPD) [26]. After “clean up” this list contains 466 binding sites upstream of 200 different genes with a little less than 30 bp in sites per intergenic region. Based on recent estimates [35] this probably corresponds to roughly a third of all the binding sites in these upstream regions. This dataset of experimentally verified sites allows us to quantitatively compare the abilities of the different algorithms to recover true binding sites using only the orthologous sequences of the five sensu stricto species as we did with the synthetic data in Figure 5. By comparing the performance on the real data with the performance on synthetic data we also learn about the effect, on the various algorithms, of the complexities in the real data that are not captured by the assumptions that the algorithms make.
For each of the 200 genes, we gathered its upstream region together with the orthologous upstream regions from S. paradoxus, S. mikatae, S. bayanus, and S. kudriavzevii. In complete analogy with the previous test on synthetic data, we ran PhyloGibbs with and without phylogeny, PhyME, EMnEM, WGibbs, and MEME on each of these 200 upstream regions. For PhyloGibbs we used Dialign [22] alignments of the upstream regions. PhyME uses the MLAGAN [29] software for its multiple alignments, and EMnEM uses ClustalW alignments [30]. Each of these algorithms was asked to search for three motifs and an expected three sites per motif (nine sites in total). The non-phylo algorithms were also asked to search for three motifs and to expect ten sites per motif. (We experimented with other site numbers, and ten gave the best overall results for the non-phylo algorithms.) For the phylo algorithms we also needed to specify the phylogenetic tree. We approximated the topology of the sensu stricto species by a star topology and set the branch lengths based on recent estimates of conservation rates at third positions in 4-fold degenerate codons between these genomes [35] as described in Materials and Methods.
The left panel of Figure 6 shows the performance of the algorithms on this dataset analogously to the left panel of Figure 5. We see that, as on the synthetic data, PhyloGibbs outperformed all other algorithms on the real data. In contrast to the performances on the synthetic data, the difference between the performances of the phylo and non-phylo algorithms was much less pronounced. At very low sensitivity, PhyloGibbs run without phylogeny performed almost equally well as PhyloGibbs with phylogeny. PhyME had high sensitivity and outperformed both EMnEM and the non-phylo algorithms at this sensitivity, but it seemed unable to make very specific predictions. EMnEM did not perform better than any of the non-phylo algorithms on these data.
Figure 6 Performance of Several Motif-Finding Algorithms on 200 Alignments of Orthologous Intergenic Regions from Five Saccharomyces Species Containing Documented Binding Sites
The left panel shows how the fraction of predicted sites that match true sites (specificity) depends on the fraction of true sites that are among the predictions (sensitivity) for PhyloGibbs (red), EMnEM (yellow), PhyME (green), PhyloGibbs without phylogeny (light blue), WGibbs (dark blue), and MEME (pink). Dashed lines correspond to one standard error. In order for the specificities, predicted by the various algorithms, to match the true specificities, we have to assume that the known sites are only a fraction of all true sites. The right panel shows what the fraction of known sites among all true sites should be in order for the algorithms' predicted specificities to match the true specificities. The black line shows an independent estimate of the fraction of real sites in these upstream regions that is documented (see text).
We believe that one important factor contributing to the smaller difference between the phylo and non-phylo algorithms is the limited reliability of the multiple alignments. Since all phylo algorithms only sample configurations consistent with the alignment, any errors in the alignment will hurt their performance. Another factor that probably plays a role is that all phylo algorithms assume that when a site occurs in a conserved block, the site must occur in all species. This is probably not always true, i.e., there are cases where only some of the sequences in an aligned block have retained the site. The non-phylo algorithms can easily deal with this by placing windows only on those sequences that have retained the site, but the phylo algorithms cannot, and a block with several binding sites may be “spoiled” by a single sequence that is missing the site.
All specificities in Figure 6 are significantly lower than those for the synthetic data in Figure 5. There are, of course, many differences between the synthetic and real data (the real data have more complex background, WMs of varying widths, varying numbers of motifs and sites, etc.), but we believe the main reason for the much lower specificity is that the specificities are based on counting only documented sites, and that many true binding sites are not yet documented. There is no reason to believe that the algorithms are more likely to recover documented binding sites than they are to recover true but undocumented binding sites. The reported specificities sr counting only documented sites thus likely underestimate the true specificities st by a factor that roughly corresponds to the fraction fd of all true sites that are documented. That is, assuming that all algorithms are equally likely to recover true but undocumented sites as they are to recover documented sites, and assuming that the algorithm's true specificity st matches the specificity sp that it predicts itself, we have
with sr the reported specificity on documented sites as shown in the left panel of Figure 6. This implied value fd is shown in the right panel of Figure 6 as a function of the predicted specificity sp. We see that PhyloGibbs predicted a fraction fd that was relatively insensitive to sp and lay between 33% and 45% over a wide range. PhyloGibbs without phylo predicted an fd in the range of 25% to 40%. Both of these predictions are consistent with the independent estimate [35] that the documented sites represent about 33% of all true binding sites (black line). As with the synthetic data, these results suggest that PhyloGibbs's own assessment of the reliability of its predictions is fairly accurate. Thus, while 18.5% of all predicted sites at a sensitivity of 50% matched documented binding sites, the true specificity is probably somewhere around 55%, and the predicted sites at sensitivity 10% are likely almost all real (A rescaled version of the left panel of Figure 6 is shown in Figure S3). In contrast, the values of fd obtained for the other algorithms were all unrealistic, for reasons that we already discussed above.
All binding sites that PhyloGibbs predicted in the upstream regions of the genes with one or more sites in SCPD are listed in Dataset S1. They can also be viewed at [36].
Inferring Yeast's “Regulatory Code”
In the previous sections PhyloGibbs inferred the locations of regulatory sites in one intergenic region at a time. Although sites for a given TF often occur in multiple copies in a single intergenic region, there are also many cases where only a single site occurs, and in those cases PhyloGibbs has to rely on conservation alone to infer the locations of the regulatory sites. However, PhyloGibbs is not limited to run on a single multiple alignment of orthologous intergenic regions but can also run on a set of multiple alignments for co-regulated genes, which should significantly increase sensitivity and specificity.
To test the performance of PhyloGibbs in this setting we used data from a recently published [27] draft transcriptional “regulatory code” of S. cerevisiae. Harbison et al. [27] performed ChIP-on-chip experiments with 203 different TFs from S. cerevisiae to identify the intergenic regions that are bound by each of them. A suite of six motif-finding algorithms was run on these intergenic regions (several algorithms also used the orthologous regions from other species), and the results were then clustered to arrive at a consensus WM for each TF. When no motif was found computationally for the intergenic regions pulled down, the literature was used, whenever possible, to define a motif. This led to predicted WMs and binding sites for 102 TFs.
We tested PhyloGibbs on the highest confidence set of intergenic regions regulated by each factor. We focused on the 45 TFs that had the fewest binding sites annotated in [27] (a minimum of three and a maximum of 25) since these are generally the most challenging to locate. For 21 of these 45 TFs, the six motif-finding algorithms employed in [27] failed to find a significant motif in the input data, and the reported motif and sites are based entirely on a consensus obtained from the literature (of the remaining 57 WMs with more than 25 or less than three annotated sites, 16 were also solely based on literature.)
We first tested whether, in contrast to the motif-finding algorithms employed in [27], PhyloGibbs was capable of recovering a significant motif that matches the literature consensus for these 21 TFs. For each TF we collected all intergenic regions that were annotated in [27] to contain at least one binding site, collected their orthologs from the other sensu stricto Saccharomyces species, produced multiple alignments using Dialign [22], and ran PhyloGibbs on each of these sets of alignments. Since each of these collections of intergenic regions will typically contain binding sites for multiple TFs, we asked PhyloGibbs to search for three motifs, with a total number of sites equaling three times the number of annotated sites for the TF in [27]. The results of this test are shown in Table 1.
Table 1 Results of PhyloGibbs on Collections of Intergenic Regions for 21 TFs for Which the Motif-Finding Algorithms in [27] Failed to Recover a Significant Motif but for Which a Literature Consensus Motif Is Available
We evaluated the results that PhyloGibbs reported for each TF in various ways. As described in Materials and Methods, PhyloGibbs reports two WMs for each motif that it finds: one constructed from the configuration C* at the end of anneal, and one by weighing the member windows of a motif with their membership probabilities obtained through tracking. We compared these WMs with the WM that is reported in the literature for each of these 21 TFs. We also compared the sites that PhyloGibbs reported, in both anneal and tracking, with the sites reported in [27]. For each TF we ran PhyloGibbs twice, first with a motif width matching the literature consensus, and then with a default width of 15. We then determined which of the motifs that PhyloGibbs reported best matches the literature motif, and we report a number of statistics for this motif (Table 1). For example, the statistics for TF GCR1 show that, for one of the motifs reported by PhyloGibbs, both the anneal WM and the tracking WM have a probability 1.0 to match the literature WM. There are ten windows with this color in the anneal configuration C*, of which six match sites annotated in [27] for GCR1. During tracking, on average 7.96 windows were members of this motif, and on average 5.17 of these members matched sites annotated in [27]. A total of seven sites were annotated for GCR1 in [27].
We see that for 16 of the 21 TFs, PhyloGibbs found a motif that matched, according to at least one statistic, the consensus motif known for this TF in the literature. PhyloGibbs thus apparently outperformed all of the motif-finding algorithms used in [27] on these 16 datasets. For the top eight TFs in Table 1 there is a good match between the WMs that PhyloGibbs reports and the literature WM as well as a significant overlap between the reported sites and the sites reported in [27]. This agreement might suggest that these groups of sites almost exhaustively capture all sites genome-wide for these eight factors. To test this we picked one example, GCR1, and compared the reported sites with known sites from the literature. In [26] there are six genes whose upstream regions have reported GCR1 sites (TPI1, CDC19, ENO2, ADH1, ENO1, and PGK1). More recently, it was shown that there are additional GCR1 sites upstream of GKL1 and GCR1 itself [37,38]. Somewhat surprisingly, of these eight genes only one (CDC19) is among the set of four upstream regions containing GCR1 annotated sites in [27]. We ran PhyloGibbs on the eight upstream regions of TPI1, CDC19, ENO2, ADH1, ENO1, PGK1, GKL1, and GCR1, and recovered a motif that perfectly matched the GCR1 literature consensus. This motif is represented with sites in all upstream regions, although the site upstream of GKL1 has a posterior probability of only 0.1. Thus, the sites PhyloGibbs found in these upstream regions are very likely true GCR1 sites. This indicates that the sites reported in [27] are far from exhaustive.
The results for PUT3 seem paradoxical. All sites PhyloGibbs reported matched sites reported in [27], but the WMs did not match. The reason for this is that the consensus for PUT3 used in [27], CGG..........CCG, has a 10-bp spacer that is presumed to contain random bases. In contrast, the motif that PhyloGibbs inferred, CGGNNNNGGNTTCCCG, is much more specific. It has been established that PUT1 and PUT2 are directly regulated by PUT3 [39]. The upstream region of PUT2 is indeed among the three upstream regions annotated with sites for PUT3 in [27], but PUT1 is not. We added this upstream region to the collection of upstream regions for PUT3 and reran PhyloGibbs. PhyloGibbs again found the PUT3 motif, which now included two good binding sites upstream of PUT1.
For MET31 the WMs matched reasonably well, and two out of three sites in configuration C* matched, but the sites did not cluster stably during tracking. According to the literature [40–42], MET31 directly regulates MET25, MET3, MET14, GSH1, and MET28. None of their upstream regions are among the upstream regions annotated with sites for MET31 in [27]. We ran PhyloGibbs on these five upstream regions and found a motif that matched the literature consensus that had sites in all five of these upstream regions that are stable under tracking. Thus, as with GCR1, all these sites are very likely true MET31 sites not annotated in [27].
For ADR1 and MAC1, both reported WMs showed a significant match to the literature motif but the reported sites overlapped only marginally with the sites reported in [27]. For ADR1 there are two genes that have known sites upstream (ADH2 and CTA1) in [26]. Neither of these have annotated sites for ADR1 in [27]. A recent microarray-based study [43] lists 74 genes as under control of ADR1, of which only PXA1 occurs among the upstream regions with sites for ADR1 in [27]. Both ADH2 and CTA1 occur in this list as well. We ran PhyloGibbs on the upstream regions of ADH2, CTA1, and PXA1 and found a motif matching the ADR1 literature consensus and containing sites in all three upstream regions. Again these sites are thus very likely true sites for ADR1 that, except for the PXA1 sites, are not contained in the annotation of [27]. For MAC1 a similar story applies. Binding sites for MAC1 are listed upstream of FRE1 and CTR1 in [26], and CTR3 and CTT1 are additionally identified as targets in the literature [44,45]. Of these only CTR1 is among the genes with sites for MAC1 in [27]. Running PhyloGibbs on these four upstream regions recovered a motif that matched the MAC1 literature consensus perfectly and that had sites upstream of FRE1 and CTR1. It also had a site upstream of orthologs of CTR3 but, curiously, not in the S. cerevisiae upstream region of CTR3, which, equally curiously, did not align well with the upstream regions of its orthologs. PhyloGibbs found no site in CTT1. This case warrants closer study.
For HAP5 and SKO1, only the anneal WM matched the literature WM. Although a reasonable number of windows occurred on average during tracking for these motifs, there was no stable core. Even the stablest window in each group was only present about 50% of the time. The membership of these groups thus fluctuated significantly during tracking, and this is reflected in the fact that the information score (see Materials and Methods) of the tracking WM is much lower than that of the anneal WM. In [46] five genes (AHP1, GLR1, GRE2, SFA1 and YML131W) are reported as direct targets of SKO1. None of their upstream regions occur among the upstream regions with annotated sites for SKO1 in [27]. We ran PhyloGibbs on the upstream regions of these five genes and found a motif that matches the literature consensus and had sites in the upstream regions of all but SFA1. Interestingly, the consensus of the motif PhyloGibbs reported, TTACGTAA, subtly differs from the literature consensus TGACGTCA. The PhyloGibbs consensus is still a palindrome but the second guanine and penultimate cysteine are replaced by thymine and adenine, respectively.
For GZF3 and RLM1 there was only a moderate match of the anneal WM to the literature WM, and no overlap whatsoever of the reported sites with the sites reported in [27]. Coffman et al. [47] report GAP1, DAL80, and UGA4 as direct target genes of GZF3. Again, none of these are among the genes annotated with sites for GZF3 in [27]. We ran PhyloGibbs on the upstream regions of these three genes and recovered a motif that significantly matched the literature motif and had sites upstream of all three genes. Interestingly, the motif that PhyloGibbs reported, GATWAGCGAT, while matching the literature consensus GATAAG, is longer and significantly more specific. Dodou et al. [48] report RLM1, SMP1, HKR1, KTR2, HSP150, and FLO1 as direct targets of RLM1. Of these, only HSP150 is among the genes with sites annotated in [27]. We ran PhyloGibbs on the set of six upstream regions from [48] and recovered a motif that reasonably matched the RLM1 literature consensus. The consensus of the motif that PhyloGibbs reported is WGCWAANNTTARAW, whereas the literature consensus is CTAWWWWTAG. PhyloGibbs found sites upstream of four (SMP1, HSP150, KTR2, and HKR1) of the six genes, with multiple sites in front of HSP150.
Finally, there were five TFs (DAL80, MOT3, ROX1, YAP6, and YOX1) for which PhyloGibbs did not find any motif matching the literature motif among the intergenic regions from [27]. DAL80 is one of a family of four GATA TFs that all bind motifs containing the consensus GATA. The experimentally best confirmed target genes for DAL80 are DAL3, DAL80 itself, GAT1, and UGA4 [47,49,50]. None of these have binding sites for DAL80 annotated in [27]. We ran PhyloGibbs on the upstream regions of these four genes and found a motif with GATAAG consensus that had at least two sites in each of the upstream regions.
Hongay et al. [51] suggest that ERG2, ERG6, and ERG9 are direct targets of MOT3. Again, none of these upstream regions have sites annotated for MOT3 in [27]. We ran PhyloGibbs on the upstream regions of these three genes but did not find a motif clearly matching the literature consensus.
Linde and Steensma [52] found 25 genes in an expression array experiment that are likely targets of ROX1, among which are the three genes (CYC1, HEM13, and ROX1) with binding sites in [26]. None of these 25 genes are among the genes with ROX1 sites annotated in [27]. We ran PhyloGibbs on the three upstream regions of CYC1, HEM13, and ROX1 and recovered a motif that perfectly matches the literature consensus and had sites in each of the three upstream regions.
For YAP6 a consensus binding site has been established by in vitro studies of different YAP proteins binding DNA [53]. As far as we can tell no clear target genes are known in the literature for YAP6.
Finally, Pramila et al. [54] report 28 target genes for the homeodomain protein YOX1. Of these only SPO12 is among the genes with sites annotated for YOX1 in [27]. YOX1 is known to interact directly with MCM1, and a motif for YOX1 was reported in [54] that was found by first identifying the MCM1 binding sites in the upstream regions of the target genes, and then searching for an additional overrepresented motif near the MCM1 sites. We also ran PhyloGibbs on the upstream regions of these 28 genes and identified a motif with consensus ATTACWTTTCCYNAW. The right end of this consensus matches the MCM1 consensus tttCC.rAt..gg, and the left end corresponds to the standard homeodomain core ATTA. This motif has sites in about half of the 28 upstream regions. Note that Pramila et al. [54] report a YOX1 consensus of yaATTa that differs from the consensus, AsAATA.TGAmr, that is reported in [27].
Table 2 shows the results of running PhyloGibbs on the remaining 24 TFs with fewer than 25 sites in [27], where their computational methods did produce a motif. The table shows that for 18 of these, both the anneal and tracking WM from PhyloGibbs matched the WM in [27], along with a significant fraction of the sites. As with the case of GCR1 above, one should not conclude from this that the reported sites in any way cover all the true sites for these TFs. For four TFs, a motif that PhyloGibbs reported had some overlap with the motif reported in [27], and there was no meaningful overlap for only two cases: SPT2 and SPT23. We examined in detail only these two cases.
Table 2 Results of PhyloGibbs on Collections of Intergenic Regions for 24 TFs for Which the Motif-Finding Algorithms in [27] Found a Significant Motif
The protein SPT2 is involved in regulation of chromatin structure and is known to interact directly with the SWI/SNF complex and with histones. SPT2 has been reported to not have any sequence specificity in its DNA binding [55], and more recently Novoseler et al. [56] have proposed that SPT2 binds to two strands of double-stranded DNA at their crossing point. Moreover, the only well-established target of SPT2 that we found in our cursory survey of the literature was the HO gene, and this gene is not among the genes annotated with sites for SPT2 in [27]. We thus believe that the motif reported in [27] is dubious. It is well established that SPT23 regulates OLE1 expression [57], but this gene is not among the genes with sites for SPT23 annotated in [27]. Given that SPT23 does not seem to have a DNA-binding domain, it is likely that SPT23 functions as a cofactor and lacks specific DNA binding on its own.
In summary, PhyloGibbs, when run on the highest quality intergenic regions and their orthologs reported in [27], found a motif that matches the literature consensus for 16 of 21 TFs, where the computational methods of [27] failed. For 11 TFs (ADR1, DAL80, GCR1, GZF3, MAC1, MET31, MOT3, RLM1, ROX1, SKO1, and YOX1), where the correspondence was weakest or nonexistent, we extracted co-regulated groups of genes from the literature. In every case there was little or no overlap between the literature list and the set of regulatory targets claimed in [27]. For all but one of the 11 (MOT3), PhyloGibbs found a motif that matched the literature consensus and reported sites in all or almost all of the upstream regions. Thus, when a solid motif was not found, the problem was likely with the set of intergenic regions in [27], not with PhyloGibbs.
Detailed comparisons of PhyloGibbs's results with the annotations of [27] are in Dataset S2. A list of all binding sites predicted by PhyloGibbs for the 45 TFs is in Dataset S3. They can also be browsed at [36].
Discussion
Motif discovery algorithms make use of a variety of different kinds of information to identify binding sites for regulatory factors in intergenic DNA. Sequence specificities for particular regulatory factors can sometimes be obtained through detailed experimentation, including DNaseI footprinting and SELEX experiments. Weight matrices representing the sequence specificities can then be used to locate putative binding sites for these regulatory factors. In this respect algorithms often look for combinations of binding sites for several WMs [58–60] that occur within a relatively short interval on the genome. Ab initio methods typically operate on sets of sequences that are thought to contain binding sites for common regulatory factors. To isolate such sets of sequences various kinds of experimental data, such as data from microarray or ChIP-on-chip studies, can be used. The algorithms then search for sequence motifs that are overrepresented among the sequences. Alternatively, ab initio methods can use orthologous sequences from related species to search for short sequence segments that appear more conserved evolutionarily than surrounding sequences, or more conserved than can be expected based on the evolutionary distances of the species.
In this paper we have presented a novel algorithm for ab initio discovery of regulatory sites that combines the search for overrepresented motifs with the analysis of sequence conservation in arbitrary collections of sequences and their orthologs. A major challenge in using orthologous sequences is distinguishing conservation due to functional constraints, such as regulatory sites, from conservation simply due to evolutionary proximity. In order to do this correctly one has to determine which sequence segments have evolved from a common ancestral segment, i.e., the sequences have to be aligned, and their phylogenetic relationships have to be taken into account. This is complicated by the fact that orthologous intergenic sequences typically cannot be trivially aligned but show a complex pattern of conserved blocks interspersed with unalignable segments. Moreover, regulatory sites are not necessarily restricted to the conserved blocks.
Focusing only on the conserved blocks, as is done in phylogenetic footprinting approaches [13–17], misses a significant fraction of true regulatory sites, and we thus chose to include all sequence. Ideally, one would sample over all possible combinations of multiple alignments and binding site configurations, but we believe that this search space is too large to search effectively. Moreover, especially for relatively closely related species, large sequence blocks can be unambiguously aligned and the search space can be significantly reduced by pre-aligning these. However, pre-aligning all orthologous sequence groups in global multiple alignments could be deleterious because global alignment often “forces” phylogenetically unrelated sequence segments together in the alignment and might introduce spurious gaps into binding sites. We thus prefer to align only those sequence segments that can be unambiguously aligned and leave the rest of the sequences unaligned. In our current implementation we use the Dialign [22] algorithm to create multiple alignments. It searches all pairs of statistically significant (over some cutoff) pairwise ungapped alignments and then, starting from the most significant, combines all mutually consistent ones into a multiple alignment. The parts of the sequences that are not part of the consistent set of pairwise alignments are left unaligned.
Recently, two algorithms [24,25] were reported that generalize the EM algorithm MEME [6] to include phylogeny. These algorithms use EM to search the space of WMs as opposed to sampling the space of binding site configurations as PhyloGibbs does. One important advantage of the latter approach is that any arbitrarily complex prior P(C) can be easily implemented, whereas the EM approaches are essentially restricted to putting an independent prior probability on each binding site occurrence. As a consequence, we have observed that the number of sites that PhyME and EMnEM predict increases or decreases dramatically as the branch lengths of the phylogenetic tree are changed, whereas PhyloGibbs's predictions are much less sensitive to the phylogeny parameters.
Another difference, also related to the prior P(C) over configurations, is in the way that multiple motifs are treated. The EM algorithms search for multiple motifs by searching for a single motif at a time, blocking its sites and iterating. In contrast, we have optimized PhyloGibbs's move set such that it can efficiently search for sites for multiple motifs in parallel. This also allows us, as we intend to do in the future, to extend PhyloGibbs's scoring function to take correlations between the positions of binding sites for different motifs into account. Finally, as shown and discussed in the results on synthetic and real data, only PhyloGibbs realistically estimates the reliability of the binding sites that it reports. The EM algorithms ignore the uncertainty associated with the WM they infer and therefore vastly overestimate the significance of the sites that they predict.
An important novel feature of our algorithm is the anneal-and-track strategy. The algorithm first uses simulated annealing to search for the configuration C* with maximal posterior probability P(C|S) and uses this as the reference configuration. In the second phase it then samples the distribution P(C|S) of all binding site configurations and compares these samples with the reference configuration C* to assign posterior probabilities to all sites it reports. This strategy allows the algorithm to assign realistic posterior probabilities to all the sites that it reports. Instead of using the annealing step, users can also specify a reference configuration C* themselves and use the algorithm to assign posterior probabilities to the motifs occurring in C* and the sites associated with them. The anneal-and-track strategy also makes the algorithm robust to prior overestimates of the number of motifs and sites that occur in the data. Superfluous sites found by the anneal will not be stably associated with a color during tracking, and superfluous motifs will have minimal membership during the tracking.
In some approaches multiple runs of one or more algorithms on the same data are used to assess motif significance. However, in order to assess which motifs recur in multiple runs, results from the different runs have to be clustered, and the only way to do this correctly is to use the same sampling method as was used to extract the motifs in the first place. Our tracking strategy circumvents the need for such post-processing of the results.
Our tests with synthetic data showed that, in the idealized situation where orthologous sequences are perfectly aligned, algorithms that take phylogeny into account significantly outperform those that do not (see Figures 3 and 5). We also showed that, given enough species and a reliable alignment, even a single site for a fuzzy motif can be accurately recovered (see Figure 4). This underscores the potential power of using orthologous sequences for regulatory site detection. It suggests that any regulatory site can be reliably recovered given an alignment of enough related species in which the given regulatory site occurs.
We used intergenic regions of S. cerevisiae that contain experimentally verified [26] binding sites, together with those of four other sensu stricto Saccharomyces species to test the extent to which different algorithms can recover regulatory sites from multiple alignments of single intergenic regions. We ran PhyloGibbs and four other motif-finding algorithms on the multiple alignments of 200 intergenic regions and showed that PhyloGibbs outperforms all other algorithms including EMnEM and PhyME, which also take phylogeny into account (see Figure 6).
We also ran PhyloGibbs on collections of intergenic region alignments of genes that were annotated in [27] to contain binding sites for a common TF based on data from ChIP-on-chip experiments. For almost all cases for which the motif-finding methods in [27] found a significant motif, PhyloGibbs reports a matching motif. More importantly, for 16 of the 21 TFs for which the six motif-finding methods in [27] failed to find a significant motif, PhyloGibbs does report a motif that matches the literature consensus. We studied in detail those TFs for which there was disagreement between PhyloGibbs's results and those in [27]. For all these TFs we found that the gene sets reported in [27] have very little overlap with targets reported in the literature. Moreover, when PhyloGibbs is run on the upstream regions of the literature targets it recovers a set of binding sites that match the literature in all but one case.
There are several issues that we intend to address in future extensions of the algorithm. First of all, we intend to extend the types and specificity of the priors that we allow. For example, when running on multiple alignments of several different upstream regions, one may sometimes have prior information that each upstream region has at least one site for a particular TF. We would thus like to allow for priors that specify not just the total number of sites, but specifically the likely number of sites in each upstream region. We also intend to extend the priors on WMs. In many cases one may already possess alignments of known sites or other specific prior information on the sequence specificity of particular WMs, and we will extend the algorithm to allow users to “seed” different colors with such specific prior information. These extensions are all straightforward to implement. A more challenging issue for the future is the improved treatment of the multiple alignment of the intergenic regions. The complex patterns of conserved blocks interspersed with unalignable sequence that is observed in multiple alignments of orthologous sequences suggests that the evolution of intergenic regions is currently not well understood. Different mechanisms that lead to insertions and deletions of various sizes, such as tandem duplications [61], probably play a significant role, and current alignment algorithms do not take such events into account. Another important facet that is currently mostly unexplored is the extent to which regulatory sites are conserved across related species. Intuitively one expects that the closer the species, the more binding sites will be shared between them, but it is currently not generally known what fraction of sites turns over as a function of evolutionary distance, and how much this varies with the TF and evolutionary lineage in question. Finally, it is clear that different binding sites have different affinities for their cognate TF, and it is conceivable that binding sites are selected in evolution not only to remain recognized by their TF, but that there is specific selection for preserving the strength of the binding site. Further investigation will be necessary to determine if realistic models of binding site evolution need to take such selection for binding site affinity into account.
Materials and Methods
WM information score.
The most useful quantity characterizing the “quality” of a WM w is its information score I:
where m is the width of the WM, b
α is the background probability of base α, and the logarithm is often calculated base 2 to express the information score in bits. Many relevant quantities regarding sets of binding sites can be expressed in terms of information scores. For instance, the fraction of random sequences of length m that “match” the WM scales as e−I. The probability that a sample of n random sequences will have base counts ni
α = nwi
α is approximately e−nI. Similarly, the likelihood ratio R of this sample of sequences stemming from a WM versus stemming from background scales as R ∝ enI. The information score thus accurately summarizes the sequence specificity of the set of binding sites that it represents.
Prior on configurations.
The simplest prior over configurations, representing “complete ignorance,” is the uniform prior, P(C) = constant, that assigns equal probability to all configurations. However this prior is “too ignorant” to work well in practice. In particular, the large majority of all configurations will consist of configurations with a very large number of windows that essentially cover the entire input data. PhyloGibbs thus allows for two kinds of priors that take into account information regarding the expected total number of sites and motifs in the dataset. First, one can run PhyloGibbs on only the subspace of configurations with a fixed number of colors c and a fixed number of total windows N (effectively setting P(C) = 0 for all configurations outside of this subspace). The second way of putting a prior on the space of configurations is by introducing an exponential prior distribution over the number of colored windows:
with n(C) indicating the number of colored windows in configuration C. For each value of p, the distribution P(C) is the maximum entropy distribution over configurations conditioned on the expected number of colored windows
. One can thus use this prior to incorporate prior knowledge of the expected total number of binding sites in the input data.
Derivation of P(Sc).
The probability P(Sc|w) that all sequences Sc in the windows belonging to color c were drawn from a particular WM w is given by
with m being the width of the WM and n
αi being the number of times that base α occurs at position i among the sequences in Sc. Since we do not know the WM, we integrate over all possible WMs (separately for each color) [28]. That is, we formally have
where P(w) is a prior distribution over the space of WMs, and the integral extends, for each position i, over the simplex w
αi ≥ 0 and
. PhyloGibbs uses Dirichlet prior distributions of the form
where the γ parameter, which is generally referred to as a pseudocount, can be set by the user (default is γ = 1). With this prior the integral can be done exactly, and we obtain
with n being the total number of sequences in color c and Γ(x) the gamma function.
Background model.
We will assume that the background sequence was generated by a Markov model of order k, where k is specified by the user and may run from zero to arbitrary length (within reason). In this model, the probability of observing a background base αi depends on the k preceding bases αi − k through αi −
1. We estimate the probabilities P(αi|αi −
1…αi − k) either from a large set of intergenic sequences provided by the user or from the set of sequences that are being sampled as follows:
where N(αi − k…αi) is the actual number of occurrences of the string αi − k…αi in either the large intergenic sequence or the input data. The pseudocount ɛ can again be set by the user. Using this model, the probability for all the uncolored sequences is
where the product is over all positions i that have color zero.
It is conceptually and computationally convenient to divide the probabilities P(S|C) by the probability P(S|C
0) of the configuration C
0 in which all windows are color zero (i.e., background). The factor P(S ∉ C|B) then cancels, and we have
For each color c the denominator can be calculated in the same way as the numerator, with background probabilities replacing the WM entries w
αi
, and with no integral.
Identifying legitimate windows.
At the start of each run, PhyloGibbs determines the set of all legitimate windows in the data. That is, it finds all locations where a window of length m can be placed, extends the window to contain sequences from all species that share aligned bases in this segment, checks for consistency of the alignment, completes the window with unaligned bases when necessary, and records all other windows that overlap this window. Formally, the procedure is as follows. All bases in all sequences are first set as “unmarked.” All bases are then examined sequence by sequence from left to right. If the current base is “ummarked,” a single-sequence window of width m is constructed starting at that base. Next, for every uppercase letter in the window all other sequences that contain an uppercase letter in that position are added until no more sequences can be added. (The expanded window will now contain, for every uppercase letter in it, all sequences where an uppercase letter occurs in that position.) The window is then examined for consistency: the relative positions of vertically aligned uppercase letters should be the same for all uppercase letters (i.e., there should be no gaps, or if gaps exist, they should extend across all sequences affected by the vertically aligned letters). A consistent window is accepted. The treatment of inconsistent windows depends on the settings of the −D option. If −D 2 is used, all inconsistent windows are simply rejected. If the −D 1 option is used, the inconsistent window is split into smaller consistent windows as follows. Recursively, the first sequence in the current window is picked and all other sequences consistent with it are collected. This process is then repeated on the remaining sequences until all sequences have been used. It is possible that the choice of splitting is dependent on the order of sequences (e.g., when three sequences are mutually inconsistent but separating any one of them renders the others consistent). Currently we ignore this complication and assume that it is uncommon in practice. Finally, the window or windows thus obtained are added to the list of available windows, and the first base for each sequence in a window is marked so that the program will not attempt to construct additional windows starting at those positions.
Derivation of P(W|w) under the evolutionary model.
Our model for the evolution of binding sites assumes that all bases mutate at a constant rate γ. When a base at position i of a binding site is mutated to letter α, we assume that the probability that selection will fix this mutation is given by the WM component w
αi. Under this simple model, the probability that a base at position i will mutate from β to α over a time t is given by [60]
where we have introduced the “proximity” q, which corresponds to the probability q = e
−γt that no mutation took place at this position during time t.
Note that as q → 0 the expression reduces to the probability w
αi of observing an independent base α at position i when sampling from the WM w. Also note that the expression has the correct composition property when an intermediate ancestor a is inserted:
To calculate the probability P(Wi|wi) of observing the set of bases Wi in column i of a window W, we next need the phylogenetic tree relating the sequences in the alignment. The phylogenetic tree of the yeast species that we used in our runs on real data is well approximated by a “star” topology, and the calculation of P(Wi|wi) simplifies significantly for this case. We thus first present the derivation for star topologies and indicate below how PhyloGibbs calculates P(Wi|wi) for more general topologies.
For a star topology, the probability P(Wi|wi) of obtaining the set of bases Wi at column i of the window W given the WM column wi is given by
where j runs over all the sequences in the window, sj is the base appearing at position i in sequence j, and qj is the proximity of sequence j to the ancestor. Since the ancestor sequence is assumed to be a sample from the WM, we assign a probability w
αi to the possibility that the ancestor had base α at position i, and sum over all possibilities α. For the whole window we of course have
The expression P(Wi|wi) is a polynomial in the WM components w
αi
, which, as can be shown with a little algebra, has N + 4 monomial terms of the form
for an alignment of N sequences. As in the previous section, we will need to integrate products of these polynomials for multiple windows, i.e., we need to calculate integrals of the form∫P(Wi|wi) P(W′i|wi)P(W′′i|wi)…dwi. While these integrals can be done analytically, the number of terms involved equals (N + 4)(N′ + 4)(N′′ + 4)…, and this quickly becomes unwieldy when the number of windows grows. Therefore, in practice we approximate the single-window polynomials P(Wi|wi) with monomials.
Approximation of WM integrals.
For each aligned window, we approximate the window column function P(Wi|wi) of equation 19 by a monomial of the form
(in this section we drop the position index i for notational simplicity). This form has the advantage that the integral over w for the product of an arbitrary number of functions of this form is straightforward:
Therefore, using this form we can easily calculate the integrals for an arbitrary number of windows.
We now choose to set the parameters c and λα such that the first moments of the distribution P(W|w) and F(λ,w,c) match. For the zeroth moment (normalization) this gives
where P(w) is a prior over the WM space. For the first moment we demand
for all α, where
and γ is the pseudocount of the prior
. Note that the first and last equality in equation 24 are always true and that our demand is the middle equality. These equations allow us to fix c and the ratios λα/λ but leave the overall scale λ still free. We use λ to approximate the second moments. We thus demand that the exact second moments
match the second moments of the approximation
as “closely” as possible. This could, for instance, be done by choosing λ such that the square-deviation is minimized. In the current implementation we set λ by, for every combination of α and β, solving for λ from the equation 〈w
α
w
β〉e = 〈w
α
w
β〉a. This yields
We then set λ equal to the average of the λs that are obtained from this equation for the 16 combinations of α and β.
In calculating the parameters λα and c, all the complicated exact polynomial integrals for the single windows need to be done. However, since we only need to do this once for every window at the start of the run, and store the results, this does not incur a significant computational cost.
Finally, it is clear that by demanding that we approximate the function P(W|w) by a monomial of the form
we are making an uncontrolled approximation. In addition, we set the λα and c by fitting the zeroth and first, and approximating the second, moments of the distribution P(W|w), and it is not clear that this is the “best” choice one can make. For instance, one could imagine fitting the λα and c such that the average square-deviation of a much larger number of moments (including moments of high order) is minimized. However, numerical experiments for a number of windows with different parameters show that, even for fairly high-order moments, in almost all cases our current approximation is quite accurate (see Table S1).
General phylogenies.
The above method of treating a star-topology phylogeny can be readily extended to deal with more general situations. A completely general phylogeny (assuming no “lateral transfer” of DNA) can be represented as a tree; the root is the last common ancestor of all given species, nodes are intermediate ancestors (last common ancestor of some, but not all, given species), and the leaves are the actual species under consideration. All unknown ancestors (root and nodes) are separately summed over. Proximities q measure the distance of leaves or nodes to the previous node, not necessarily the last common ancestor.
Consider such a phylogenetic tree that does not have a star topology (i.e., contains internal nodes other than the root). At least one of the intermediate nodes must be such that all its children are leaves. Let the unknown base for this node at column i be β (summed over), and let the base for its parent node (again, not necessarily the root) be α. The subtree below β contributes a factor P
β to the total probability, given by
where the full expression would contain other factors involving α as well as a sum over α; q
β is the proximity of β to its immediate ancestor α, the product runs over children of β indexed by n, sn is the base of the nth descendant, and qn is the proximity of the nth descendant to β. T is given by equation 17. In particular,
Substituting this into equation 28, we get two terms:
The first term simply removes the node β and attaches all its children to α (with unchanged proximities). The second term—identical, apart from a prefactor, to equation 19—can be treated as an independent factor to anything it multiplies, completely decoupled from the sums over α and other ancestors. In other words, with probability q
β, base β is the same as α and all its leaves can be attached directly to α, and with probability 1 − q
β, base β is mutated from α and can be treated as a new, independent ancestor for all of its descendants, disconnected from the rest of the tree.
By repeating this process, one can reduce any tree to a sum of products of star-phylogeny subtrees with appropriate prefactors. PhyloGibbs then applies the monomial approximation described in the previous section to each of the star-phylogeny subtrees, as well as to the final sum. Note, however, that the number of terms involved may grow exponentially with the number of species. As the number of species becomes large we thus need to make additional approximations to make this procedure computationally feasible.
Move set.
A single time step of the algorithm consists of a “cycle” of a fixed number of moves of each of the types outlined in the following paragraphs.
Window-shift moves preserve the total number of colors, and the total number of colored windows (but may redistribute the windows among existing colors). We choose one of the presently colored windows at random. If it is the only one in its color, we make no operation (but to ensure detailed balance we update the time counter by one). If it is not the only window in its color, we color it zero (i.e., deselect it), and choose a new window from all of the available color-zero windows (including the window we selected) to replace it. The new window can have any of the existing colors, not necessarily the same as the window it is replacing. This move is computationally expensive, since if there are N available windows and c available colors, we have to calculate the scores for Nc potential moves, but it allows for rapid convergence.
Color-change moves allow for changes in the number of windows and the number of colors, while satisfying detailed balance. We select any of the existing windows, including color-zero windows. If the chosen window overlaps a non-zero-colored window then this window is blocked and we make no operation (but update the time counter). Otherwise, we reassign a color to the window, which may be zero, one of the existing colors, or a new color. Note that if the window was the only one in its color, a “new color” means “the same color as before.” The window-shift moves are not ergodic by themselves because they stay inside a subspace of fixed N and c (respectively, number of colored windows and number of colors). The color-change moves are ergodic but do not have good convergence properties, so an alternation with window-shift moves is desirable.
With the previous two moves it is possible for the sampler to get stuck in a local optimum where the windows in a given color are all shifted by an equal amount from their best location. The global shift move addresses this problem. This move picks a color at random, and samples all ways of coherently shifting every window in that color by a fixed amount without “colliding” with an already-colored window.
Maskbit-flip moves are the final move type. Long motifs tend to be fuzzy, and not every position is sharply defined. Sometimes, the score of a collection of sites can be improved by scoring a subset of its columns according to the background model rather than assuming they derive from a WM. We thus allow the “masking” of certain columns, comparing whether or not the overall score is improved by scoring them according to background. For each color we maintain a mask, and sample over the states (zero or one) of the mask bits. In our experience, allowing such masking can increase performance for long motifs that contain nonconstrained sequence, such as occur in bacteria when TFs bind as dimers. However, for short motifs the enlargement of the configuration space that is associated with these masks may result in poorer discrimination.
Tracking.
After each cycle during the tracking phase, the best-matching color c̃ from the current configuration C is found for each color c of the reference configuration C*. To do this we define a match M(c,c̃) between colors c and c̃ as follows. For all shifts −m/2 ≤ s ≤ m/2 (the window width is m), we shift all the plus-strand windows in c̃ by s to the right and all the minus-strand windows in c̃ by s to the left. Note that, since different parts of the multiple alignments will span different subsets of species, we have to define carefully what we mean by the word “shift.” We only consider two windows X and Y shifted versions of each other if they both span the same set of species, and if the position of the start of the window is shifted by the same amount in each of the sequences in the windows. Thus, in general, shifted versions will only exist for some of the windows in c̃.
After shifting all the windows in c̃ by an amount s, we then count the number of shifted windows n(s) that exactly match a window in c. The score M(c,c̃) is given by
Note that this corresponds to the maximal amount of overlap between the sites in c and the sites in c̃ when counting only sites in c̃ that are shifted by a common amount with respect to c. We also keep track of the shift s at which the maximum occurred. Once we have determined the color c̃ that maximizes M(c,c̃) and the shift s at which the maximum occurred, we then add a count of one to n(w,c) for every window w that is obtained by shifting the windows in c̃ by s. The counts n(w,c) record the number of times that each window w is associated with reference color c. At the end of tracking we divide the counts n(w,c) by the total number of cycles n to obtain p(w,c) = n(w,c)/n. For each reference color c PhyloGibbs reports all windows w for which p(w,c) ≥ p
min sorted from large to small p(w,c). These lists of membership probabilities provide a summary of the distribution P(C|D) using C* as a reference configuration.
Synthetic data runs.
For Figure 3 we generated random intergenic ancestral sequences of length L = 500 with s = 4 sites sampled from a single WM. For the different panels in Figure 3 we used WMs of width m = 10 with polarizations p = 0.6, p = 0.75, and p = 0.9 and random WMs for the lower right panel. For the data in each panel we generated gapless alignments of S = 5 descendant sequences at proximities q = 0.1 through q = 0.9 in steps of 0.1. For each parameter setting we generated 50 different synthetic datasets. PhyloGibbs with phylogeny was asked to search for four multi-sequence windows for a single WM of width ten, to assume the correct proximity q for all species, and to use a background model where each base occurs with probability 1/4. The non-phylo algorithms treat the five input sequences as independent and were asked to search for 20 single-sequence sites for a WM of width ten. The precise command lines employed were as follows: for PhyloGibbs with phylogeny, –D 1 –G q –m 10 –I 4 –f inputfile –N −1; for PhyloGibbs without phylogeny, –D 0 –m 10 –I 20 –f inputfile –N −1; for WGibbs, –PBernoulli inputfile 10 20 –Z –n; and for MEME, inputfile –dna –mod anr –w 10 –nsites 20 –nmotifs 1 –nomatrim –revcomp –maxiter 500 –text –nostatus.
As a measure of performance we took the fraction of all the bases in real sites that overlapped predicted sites and averaged it over all datasets for each parameter setting. This “overlap” thus runs from zero to one. For each parameter setting the standard error of the overlap is given by
where N is the number of datasets, oi is the overlap for dataset i, and 〈o〉 is the average overlap.
For Figure 4 we generated random ancestor sequences of length L = 500, embedded a single site for a random WM of width ten, and created S descendant sequences of proximity q = 0.5 for S running from two to 30. For each value of S we created 250 datasets and ran PhyloGibbs with the following command line settings: –D 1 –G 0.5 –m 10 –I 1 –N −1 –f inputfile. For each S we calculated the average overlap 〈o〉 and standard error as in equation 32.
The data for Figure 5 were generated, for each of 250 intergenic regions, by picking three random WMs of width m = 10, sampling three sites from each, embedding them in a random ancestral sequence of length 750, and creating five descendant sequences at q = 0.5. The phylo algorithms were asked to find sites for three WMs of width ten, with an expected number of three (multi-species) sites per WM. They were all given the correct phylogenetic tree (with star topology) and (gapless) multiple alignment. The non-phylo algorithms treat the five descendant sequences from each synthetic intergenic region as independent and so the expected number of sites per motif was set to 15 for these. The command-lines employed were as follows: for PhyloGibbs with phylogeny, –D 1 –G 0.5 –m 10 –I 3,3,3 –N −1 –f inputfile; for PhyME, –N 1 infile blkfile –w 10 –nmotifs 3 –revcompW –ot 0.05 –nsites 3 –niter 50 –nseediter 10 –K 5 –pf phylogenytree.txt –tree; for PhyloGibbs without phylogeny, –D 0 –m 10 –I 15,15,15 –N −1 –f inputfile; for WGibbs, –PBernoulli inputfile 10,10,10 15,15,15 –F –Z –n –i 1500 –S 30 –p 60; and for MEME, –dna –mod anr –w 10 –nsites 15 –nmotifs 3 –revcomp –nomatrim –maxiter 500 –nostatus. EMnEM uses a control file. The relevant parameters that we set were as follows: –w 10 –e 3.0 –b 1 –n 3 –u 1 –r 0 –m 0.
To produce the left panel of Figure 5 we recovered, for each algorithm, all the predicted sites and sorted them by their posterior probability. We then obtained a series of sublists li by excluding all predicted sites below a cutoff posterior probability ci. We chose the cutoffs ci such that at c
0 all predicted sites were included in l
0, at c
1 all but the last 100 sites were included in l
1, and generally li had all but the bottom 100i predicted sites. For each list li we then calculated the number of bases A in all intergenic regions that were hit by sites in this list, the total number of bases T in true (planted) sites, and the number of bases I in the intersection of these two sets. Given these counts, the sensitivity is given by I/T and the specificity is given by
and the standard error we similarly estimate as
This estimate of standard error correctly takes into account the fact that as the number of predictions A gets larger, our estimate of the algorithm's specificity becomes more precise. The factors ten and 0.1 are there to (approximately) take into account that the A predicted bases are not all mutually independent but that they come in windows of m = 10 consecutive bases.
For the right panel of Figure 5 we used the same estimates of the specificities and their standard errors, but plotted these as a function of the specificity spi that the algorithm predicts for each sublist li. This predicted specificity spi for the list li is obtained by averaging the posterior probabilities of all the predicted sites in list li.
Yeast data runs.
We “cleaned up” the dataset of experimentally documented binding sites from the SCPD [26] as follows. In its original form it contained 726 binding sites regulating 234 different genes. We removed sites that either lay in coding regions or that lay more than 1,000 bp upstream from translation start, and fused the overlapping sites that remained. After this we were left with 466 sites for 200 genes.
The upstream regions of the 200 S. cerevisiae genes and their orthologs were obtained from the Saccharomyces Genome Database [62]. For the sequence of S. paradoxus we used data from the MIT group [16], while for S. bayanus, S. mikatae, and S. kudriavzevii we used data from the Washington University group [15]. For each group of orthologs we took either the entire intergenic region up to the neighboring coding sequences, or 1,000 bp, whichever was shorter. Not all genes in our set of 200 have orthologs assigned in all other species. There were a total of 796 intergenic regions from five species for our 200 genes. This means that on average we had four orthologs per gene. We did not cross-check the accuracy of the annotation of orthologous genes in the downloaded files. Instead, we performed the following simple test: we flagged a 5′-UTR sequence as “dubious” if it had fewer than 100 bases or if, after aligning with Dialign, fewer than one in ten bases were marked as “aligned” (i.e., in capitals) with other sequences. Only 20 out of these 796 sequences got flagged by one of these criteria; we therefore believe that over 97% of the orthologous sequences are likely accurate, and we simply retained all in our runs.
We aligned each set of orthologous intergenic regions with Dialign [22] using the following command line: dialign2–2 –n –thr 5 –fa infile, where the setting –thr 5 ensures that only significant blocks will be aligned. We discovered, however, that the current implementation of Dialign (version 2.2.1) has a severe bug in the way it formats and displays its output. Unrelated sequence segments are sometimes reported in such a way that it appears they are aligned. For example, if one feeds Dialign four sequences, two of which contain all adenines, and two all cysteines, then Dialign will appear to align all of these together without gaps, as opposed to two blocks of pairs (the authors have been informed privately). The bug is in the assembly of fragments to the output file, and we used a wrapper, written by Michael Mwangi, to correctly assemble the fragments. Meanwhile we developed our own multiple-alignment algorithm, Sigma, which was not yet used for the results reported in this paper and which will be described elsewhere (R. Siddharthan, unpublished data).
For PhyloGibbs, EMnEM, and PhyME we needed to specify the phylogeny of the Saccharomyces species. The topology of the species tree can be determined unambiguously [63]. It has S. cerevisiae on one end and the other species branching off from it in the following order (from near to far): S. paradoxus, S. mikatae, S. kudriavzevii, and S. bayanus. We approximated this tree by a star topology. For PhyloGibbs we needed, for each of the species i, the probability qi that since the common ancestor no mutation took place in any given base. These “no mutation” probabilities, which we call “proximities,” can be best estimated by looking at the conservation statistics of “neutral” positions in the genome. It was recently shown [35] that conservation rates between S. cerevisiae and the other species at third positions of 4-fold degenerate codons are approximately constant across the genome. The conservation rates c reported in [35] are c
cer,par = 0.74, c
cer,mik = 0.6, and c
cer,bay = 0.52. In the approximation of a star topology, the conservation rates ci,j are given in terms of the proximities qi and qj through
Assuming that q
cer = q
par we obtain q
cer = q
par = 0.8, q
mik = 0.58, and q
bay = 0.45. No conservation rate was reported in [35] for S. kudriavzevii. From the topology, proximity q
kud should lie between those of S. mikatae and S. bayanus and we simply set it to q
kud = 0.5. PhyloGibbs, PhyME, and EMnEM were all run with this phylogenetic tree. EMnEM requires branch lengths in terms of the number of substitutions per site, and we used q = e
−n to determine the number of substitutions n in terms of the proximity q.
For reference we again give the command lines that we used in running the algorithms on the 200 genes with documented sites. For PhyloGibbs with phylogeny we used –D 1 –T 0.35 –m 10 –N 3 –F bgfile –I 3,3,3 –E 0.01 –f infile –L (cer:0.8,par:0.8,mik:0.58,kud:0.5, bay:0.45). Here bgfile is a fasta file with all S. cerevisiae intergenic sequences from which the background model is constructed. The setting –T 0.35 sets the pseudocount of the WM prior to 0.35 to account for the fact that TFs in S. cerevisiae generally have higher information scores than random WMs. We ensured that the fasta header for each sequence identified the name of the species from which it derived. Finally, the setting –E 0.01 instructs PhyloGibbs to report sites with probabilities as low as 0.01 (instead of the default 0.05). For EMnEM the relevant parameters in the control file were –w 10 –e 3.0 –t 0.05 –b 1 –n 3 –u 1 –r 0 –m 0. EMnEM was also provided with the phylogenetic tree as described above. It uses ClustalW alignments [30] of the upstream regions. For PhyME we used –N 1 infile blkfile –w 10 –nmotifs 3 –revcompW –ot 0.05 –nsites 3 –niter 50 –nseediter 10 –b –K 5 –pf phylogenytree.txt –tree. PhyME uses MLAGAN [29] for its alignments and parses these in its own specific way. The results of this parse are in the blkfile. For PhyloGibbs without phylogeny we used –D 0 –T 0.35 –m 10 –N 3 –F bgfile –I 10,10,10 –E 0.01 –f infile; for WGibbs, –PBernoulli infile 10,10,10 10,10,10 –Z –n; and for MEME, infile –dna –mod tcm –w 10 –nmotifs 3 –wg 1000000 –nomatrim –maxiter 500 –maxsize 1000000 –revcomp –bfile bgfile.
We should point out that the performances of the different algorithms may vary as one varies parameter settings. We experimented with different parameter settings for each of the algorithms but none substantially changed the results shown in Figure 6. For all parameter settings that we tested, PhyloGibbs with phylogeny outperformed all other algorithms. We did notice that PhyME and EMnEM were much more sensitive to the phylogenetic tree than PhyloGibbs was. PhyME showed best performance with the tree that we show here. EMnEM could be made to perform better than the non-phylo algorithms by using a tree with shorter branch lengths.
The specificity-versus-sensitivity plots in the left panel of Figure 6 were obtained almost identically to what was described for the synthetic data. The only difference was that, instead of counting the precise number of bases in predicted sites overlapping bases in true sites, we considered any predicted site to “hit” a true site if it overlapped the true site by at least 5 bp (half of the predicted site's length). We did this because the precise extent of the known sites seems often poorly defined: Typically one finds that different sites that are annotated for the same TF in [26] can have widths that vary substantially.
For the right panel of Figure 6 we show, as in the right panel of Figure 5, the predicted specificities sp of the different sublists on the horizontal axis, but on the vertical axis we show the ratio sr/sp between the measured specificity sr on documented sites and the specificity sp that the algorithm predicts.
Regulatory code.
We used version 24 of the regulatory code from [27]. In particular we used the set of “final motifs,” and the highest-confidence binding sites (binding with p < 0.001 and conserved in at least two other yeasts) based on these motifs. The former consists of a file (Final_InTableS2_v24.motifs) with 102 WMs for 102 TFs, and the latter consists of a file (IGR_v24.3.p001b.GFF) with the genomic locations of 3,353 predicted binding sites for these 102 WMs. From the set of WMs we selected the 45 WMs for which there were at least three annotated sites and at most 25. The file Final_InTableS2_v24.motifs notes which WMs are based solely on sites/consensus reported in the literature. Among the 45 WMs that we selected there were 21 that were literature based, i.e., no significant motif was found by the computational methods employed in [27]—these are the WMs shown in Table 1. For each of the 45 selected WMs we collected all intergenic regions with annotated binding sites and their orthologs in the four other sensu stricto species, and aligned them with Dialign as described above. We then ran PhyloGibbs twice on each of the 45 collections of intergenic regions. For a WM with a total of S annotated sites in [27] and a motif width of l in [27] we used the following command line settings for the two runs: (1) –D 1 –L (cer:0.8, par:0.8, mik:0.58, kud:0.5, bay:0.45) –T 0.25 –m l –N 3 –F bgfile –I S,S,S –f infile and (2) –D 1 –L (cer:0.8, par:0.8, mik:0.58, kud:0.5, bay:0.45) –T 0.25 –m 15 –N 3 –F bgfile –I S,S,S –f infile. That is, we used both the annotated binding site width and a default width of 15. Since there generally are binding sites for multiple WMs in the set of intergenic regions, we let PhyloGibbs search for three different motifs with a total number of sites equaling three times the number of annotated sites in [27].
To compare the results of PhyloGibbs with those of [27] we compared the configurations of binding sites that PhyloGibbs reported with all the motifs reported in [27]. In particular, for each motif that PhyloGibbs reports it outputs two alignments of predicted sites. One alignment consists of the sequences that have a common color in the reference configuration C*. The other consists of the time-averaged alignment of sequences that associate with this reference color during tracking. For each WM w in the file Final_InTableS2_v24.motifs we multiplied the WM entries w
αi by the total number of sites S annotated for the WM to obtain an alignment m of all the binding sites, with m
αi = w
αi
S the number of sites that have base α at position i. For each pair of one such alignment from [27] and a reported alignment from PhyloGibbs we calculated the probability that both alignments were drawn from a common WM.
Let n be one of the alignments of sites reported by PhyloGibbs and m be an alignment of sites from [27], with n
αi and m
αi being the number of times base α occurs at position i of alignments n and m, respectively. We now calculate the probability that these two alignment were sampled from a common WM, taking into account the possibility that n and m may be shifted or reverse-complemented with respect to each other. Assume n has width ln and m has width lm and assume an alignment of n and m in which n is shifted k positions to the right with respect to m. The total number of times t
αi that base α occurs at position i in this joint alignment is given by (1) t
αi = m
αi when 1 ≤ i ≤ k, (2) t
αi = m
αi + n
α(i − k) when (k + 1) ≤ i ≤ lm, and (3) t
αi = n
α(i − k) when lm + 1 ≤ i ≤ ln + k. The probability to draw this joint alignment t from a WM is
where γ is the pseudocount of the prior over WM space and ti is the total number of bases in column i of the joint alignment t. We here use the uniform prior γ = 1. The probability to draw n and m from two separate WMs is similarly given by P(n)P(m) with each factor given by the same equation 36. Thus, the posterior probability P(t|n,m) that n and m, forming joint alignment t, were drawn from a common WM is given by
where π is the prior probability. For each alignment n that PhyloGibbs reported there were 102 alignments m from [27] (one for each TF), and for each of those we considered all relative shifts and reverse-complement combinations in which at least 4 bp overlapped between m and n. For alignments of length ln and lm there are 2(ln + lm − 8) such combinations. We set π such that the prior probability was 1/2 that any of these shift/strand combinations from any of the 102 WMs m gave an alignment t that was sampled from a common WM. That is, we set
Finally, for each combination n and m we calculated the maximum of P(t|n,m) over all 2(ln + lm − 8) shift/strand combinations t to obtain
In the Dataset S2 we show all P(n,m) that are larger than 1/4.
In addition, for each combination of a reported motif from PhyloGibbs and a TF with annotated sites in [27] we calculated the fraction of sites in the motif that overlapped (by at least 4 bp) a site annotated for that TF in [27]. For the motifs reported in tracking we again weighed each site by its posterior probability in calculating this fraction. Dataset S2 shows all combinations of reported motifs and TFs for which this fraction is non-zero.
The results in Tables 1 and 2 were calculated as follows. After running PhyloGibbs on the set of upstream region alignments for one of the 45 TFs, we analyzed all three motifs that PhyloGibbs reported and determined which one best matched the motif m reported for the TF in [27]. For each motif we obtained the alignment of sequences nr reported in the reference configuration C* and the time-averaged alignment nt obtained for this motif through tracking, and calculated the probabilities P(nr ,m) and P(nt ,m) that these alignments were sampled from the same WM as the alignment m of sites reported in [27]. We also calculated the fractions f(nr ,m) and f(nt ,m) of sites in nr and nt that overlapped sites annotated for the motif m in [27]. The total score s(n) of the motif was simply given by the sum s(n) = P(nr ,m) + P(nt ,m) + f(nr ,m) + f(nt ,m). For the motif (out of three) that maximizes s(n), Tables 1 and 2 show P(nr ,m) in the second column and P(nt ,m) in the third column. The fourth column shows the total number of sites |nr| in this motif (color) in reference configuration C* and the total number of those f(nr ,m)|nr| that overlap sites annotated for m in [27]. The fifth column shows the same statistics for the tracked set of sites nt, where again each site is weighed by its posterior probability. That is, |nt| is the sum of the posterior probabilities of the sites in this motif. Finally, the sixth column shows the total number of sites |m| that were annotated for motif m in [27].
For 11 TFs we gathered sets of target genes from the literature, collected their orthologs from the other sensu stricto species, obtained multiple alignments with Dialign, and ran PhyloGibbs on these sets of multiple alignments. The following command line options were used for all these runs: –D 1 –T 0.25 –a 300 –S 300 –L (cer:0.8, par:0.8, mik:0.58, kud:0.5, bay:0.45) –N 3 –F bgfile –f infile. A summary of the results of these runs, and the remaining parameter settings used, are shown in Table 3.
Table 3 Results of PhyloGibbs on Multiple Alignments of Upstream Regions Taken from the Literature
Detailed results, and the locations of all the binding sites newly identified in these runs, can be found in Datasets S4 and S5.
Supporting Information
Dataset S1 Predicted Sites on Genes with Sites in SCPD
This file lists all sites with posterior probability 0.05 or higher that PhyloGibbs predicted on the upstream regions of the genes that have one or more binding sites annotated in the SCPD [26]. The sites are ordered first by posterior probability, then by the name of the open reading frame (ORF), and finally by the number of the motif in which the site occurred. An example line from the file is “YPR191W (−175, −166) rev 0.97 taagaCGGGGCGGGCcttct 3.” The first column shows the name of the ORF, the second column shows the location of the site relative to the ATG of the ORF, the third column shows the strand on which the site occurs, the fourth column shows the posterior probability of the site, and the fifth column shows the sequence of the site (in capitals) plus five bases to the left and right of the site.
(141 KB TXT)
Click here for additional data file.
Dataset S2 Comparison of the PhyloGibbs Predictions with Those of Harbison et al. [27]
This file summarizes the comparisons of the results of PhyloGibbs on the data from [27] with those reported in [27].
(99 KB TXT)
Click here for additional data file.
Dataset S3 All Predicted Sites on the Data from [27]
This file contains all binding sites with posterior probability at least 0.05 that PhyloGibbs predicted for the 45 TFs with between three and 25 sites annotated in [27]. For each TF PhyloGibbs was run on the five-species multiple alignments of all upstream regions with sites annotated in [27] and asked to predict three motifs. In this file we show only the predictions for the motif that best matched the motif reported in [27]. The format of the lines is very similar to that of the lines in Dataset S1. Example for a predicted site for TF ADR1: ADR1 YKL016C (−388,–383) fwd 0.58 tacTCCAATatt harb_lit. The site occurs 388 to 383 bases upstream of the ATG of the ORF YKL016C. It occurs on the forward strand of the genome and has a posterior probability 0.58. The fifth column shows the sequence of the site in capitals plus half a site length to the left and right. Finally, the last column shows if [27] found the WM of this TF by computational means or if they simply copied the motif reported in the literature.
(42 KB TXT)
Click here for additional data file.
Dataset S4 Co-Regulated Gene Sets Gathered from the Literature
For 11 TFs we gathered lists of genes that are known to be regulated by the TF from the literature. This file gives the list of ORF names of these genes for each of the 11 TFs. Example: DAL80 YKR034W YIR032C YDL210W YFL021W. This line shows that the TF DAL80 is reported in the literature to regulate the ORFs YKR034W, YIR032C, YDL210W, and YFL021W.
(10 KB TXT)
Click here for additional data file.
Dataset S5 Predicted Sites for the Literature Gene Sets
This file has the same format as Dataset S3 and shows all predicted sites for the 11 TFs in the upstream regions of the genes in Dataset S4. Example: GCR1 YAL038W (−263, −256) rev 0.81 ttttAGGAAGACacta. This example shows a predicted site for TF GCR1 which occurs from 263 to 256 bases upstream of the ATG of YAL038W, occurs on the negative strand, and has posterior probability 0.81.
(9 KB TXT)
Click here for additional data file.
Figure S1 Analog of Figure 3 with Yeast WMs and Proximities
Results analogous to those shown in Figure 3 but with “real” WMs representing known binding specificities of yeast TFs, and using a phylogenetic tree with branch lengths proportional of those of the Saccharomyces sensu stricto species.
(62 KB PDF)
Click here for additional data file.
Figure S2 Analog of Figure 5 with Yeast WMs and Proximities
Results analogous to those shown in Figure 5 but with “real” WMs representing known binding specificities of yeast TFs, and using the phylogenetic tree of the Saccharomyces sensu stricto species.
(110 KB PDF)
Click here for additional data file.
Figure S3 Rescaled Specificity/Sensitivity of the Predictions on SCPD Genes
Results as in the left panel of Figure 6 but with specificities rescaled assuming that only 40% of all true binding sites are documented.
(60 KB PDF)
Click here for additional data file.
Table S1 Accuracy of the WM Polynomial Approximation
This table shows a comparison of the exact WM integrals with the monomial approximation that our algorithm employs.
(7 KB PDF)
Click here for additional data file.
Support was provided from the National Science Foundation, grant DMR-0129848. EV received support from the Swiss National Science Foundation, project 3152A0–105972. Michael Mwangi programmed the script to reformat the Dialign output. Nicolas Buchler supplied 5′-UTR yeast sequences, based on the publicly available ones, from which coding shadows had been removed. RS thanks the Indian Lattice Gauge Theory Initiative for computer time on the “Kabru” cluster at the Institute of Mathematical Sciences. EV thanks Saurabh Sinha for help running PhyME and for useful comments on the manuscript.
Competing interests. The authors have declared that no competing interests exist.
Author contributions. RS, EDS, and EvN conceived and designed the experiments. RS and EvN performed the experiments and analyzed the data. RS, EDS, and EvN contributed reagents/materials/analysis tools and wrote the paper.
A previous version of this article appeared as an Early Online Release on October 27, 2005 (DOI: 10.1371/journal.pcbi.0010067.eor).
Abbreviations
EMexpectation maximization
ORFopen reading frame
SCPDPromoter Database of Saccharomyces cerevisiae
TFtranscription factor
WMweight matrix
==== Refs
References
Berg OG von Hippel PH 1987 Selection of DNA binding sites by regulatory proteins: Statistical-mechanical theory and application to operators and promoters J Mol Biol 193 723 750 3612791
Durbin R Eddy S Krogh G Mitchison G 1998 Biological sequence analysis Cambridge University Press 356 p.
Djordjevic M Sengupta AM Shraiman BI 2003 A biophysical approach to transcription factor binding site discovery Genome Res 13 2381 2390 14597652
Lawrence CE Altschul SF Boguski MS Liu JS Neuwald AF 1993 Detecting subtle sequence signals: A Gibbs sampling strategy for multiple alignment Science 262 208 214 8211139
Liu JS Neuwald AF Lawrence CE 1999 Markovian structures in biological sequence alignment J Am Stat Assoc 94 1 15
Bailey T Elkan C 1994 Fitting a mixture model by expectation maximization to discover motifs in biopolymers Proc Int Conf Intell Syst Mol Biol 2 28 36 7584402
Stormo GD Hartzell GW 1989 Identifying protein-binding sites from unaligned DNA fragments Proc Natl Acad Sci U S A 86 1183 1187 2919167
Hughes JD Estep PW Tavazoie S Church GM 2000 Computational identification of cis-regulatory elements associated with groups of functionally related genes in Saccharomyces cerevisiae
J Mol Biol 296 1205 1214 10698627
Liu X Liu JS Brutlag DL 2001 Bioprospector: Discovering conserved DNA motifs in upstream regulatory regions of co-expressed genes Pac Symp Biocomp 127–138
Liu XS Brutlag DL Liu JS 2002 An algorithm for finding protein-DNA binding sites with applications to chromatin immmunoprecipitation experiments Nat Biotechnol 20 835 839 12101404
McCue LA Thompson W Carmack CS Ryan MP Liu JS 2001 Phylogenetic footprinting of transcription factor binding sites in proteobacterial genomes Nucleic Acids Res 29 774 782 11160901
McCue LA Thompson W Carmack CS Lawrence CE 2002 Factors influencing the identification of transcription factor binding sites by cross-species comparison Genome Res 12 1523 1532 12368244
Blanchette M Schwikowski B Tompa M 2002 Algorithms for phylogenetic footprinting J Comput Biol 9 211 223 12015878
Blanchette M Tompa M 2002 Discovery of regulatory elements by a computational method for phylogenetic footprinting Genome Res 12 739 748 11997340
Cliften P Sudarsanam P Desikan A Fulton L Fulton B 2003 Finding functional features in Saccharomyces genomes by phylogenetic footprinting Science 301 71 76 12775844
Kellis M Patterson N Endrizzi M Birren B Lander ES 2003 Sequencing and comparison of yeast species to identify genes and regulatory elements Nature 423 241 254 12748633
Rajewsky N Socci ND Zapotocky M Siggia ED 2002 The evolution of DNA regulatory regions for proteogamma bacteria by interspecies comparisons Genome Res 12 298 308 11827949
Dermitzakis ET Bergman CM Clark AG 2003 Tracing the evolutionary history of Drosophila regulatory regions with models that identify transcription factor binding sites Mol Biol Evol 20 703 714 12679540
Emberly E Rajewsky N Siggia ED 2003 Conservation of regulatory elements between two species of Drosophila
BMC Bioinformatics 4 57 14629780
Siggia ED 2005 Computational methods for transcriptional regulation Curr Opin Genet Dev 15 214 221 15797205
Siddharthan R van Nimwegen E Siggia ED 2005 PhyloGibbs: A Gibbs sampler incorporating phylogenetic information Eskin E Workman C RECOMB 2004 satellite workshop on regulatory genomics Berlin Springer-Verlag 30 41
Morgenstern B Dress A Werner T 1996 Multiple DNA and protein sequence alignment based on segment-to-segment comparison Proc Natl Acad Sci U S A 93 12098 12103 8901539
Wang T Stormo G 2003 Combining phylogenetic data with co-regulated genes to identify regulatory motifs Bioinformatics 19 2369 2380 14668220
Moses AM Chiang DY Eisen MB 2004 Phylogenetic motif detection by expectation-maximization on evolutionary mixtures Pac Symp Biocomput 324–335
Sinha S Blanchette M Tompa M 2004 PhyME: A probabilistic algorithm for finding motifs in sets of orthologous sequences BMC Bioinformatics 5 170 15511292
Zhu J Zhang MQ 1999 SCPD: A promoter database of the yeast Saccharomyces cerevisiae
Bioinformatics 15 607 611 10487868
Harbison CT Gordon DB Lee TI Rinaldi NJ Macisaac KD 2004 Transcriptional regulatory code of a eukaryotic genome Nature 431 99 104 15343339
van Nimwegen E Zavolan M Rajewsky N Siggia ED 2002 Probabilistic clustering of sequences: Inferring new bacterial regulons by comparative genomics Proc Natl Acad Sci U S A 99 7323 7328 12032281
Brudno M Do CB Cooper GM Kim MF Davydov E 2003 LAGAN and Multi-LAGAN: Efficient tools for large-scale multiple alignment of genomic DNA Genome Res 13 731 731
Thompson JD Higgins DG Gibson TJ 1994 CLUSTAL W: Improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice Nucleic Acids Res 22 4673 4680 7984417
Bray N Pachter L 2004 MAVID: Constrained ancestral alignment of multiple sequences Genome Res 14 693 699 15060012
Pollard DA Bergman CM Stoye J Celniker SE Eisen MB 2004 Benchmarking tools for the alignment of functional noncoding dna BMC Bioinformatics 5
Blanchette M Kent WJ Riemer C Elnitski L Smit AF 2004 Aligning multiple genomic sequences with the threaded blockset aligner Genome Res 14 708 715 15060014
Liu JS 2001 Monte Carlo strategies in scientific computing Springer-Verlag 360 p.
Chin CS Chuang J Li H 2005 Genome-wide regulatory complexity in yeast promoters: Separation of functionally conserved and neutral sequence Genome Res 15 205 213 15653830
Swiss Institute of Bioinformatics, Biozentrum Universität Basel 2005 SwissRegulon [database] Available: http://www.swissregulon.unibas.ch . Accessed 7 November 2005.
Herrero P Flores L de la Cera T Moreno F 1999 Functional characterization of transcriptional regulatory elements in the upstream region of the yeast GLK1 gene Biochem J 343 319 325 10510295
Sasaki H Kishimoto T Mizuno T Shinzato T Uemura H 2005 Expression of GCR1, the transcriptional activator of glycolytic enzyme genes in the yeast Saccharomyces cerevisiae, is positively autoregulated by Gcr1p Yeast 4 305 319
Axelrod JD Majors J Brandriss MC 1991 Proline-independent binding of PUT3 transcriptional activator protein detected by footprinting in vivo Mol Cell Biol 11 564 567 1986247
Blaiseau PL Isnard AD Surdin-Kerjan Y Thomas D 1997 Met31p and Met32p, two related zinc finger proteins, are involved in transcriptional regulation of yeast sulfur amino acid metabolism Mol Cell Biol 17 3640 3648 9199298
Blaiseau PL Thomas D 1998 Multiple transcriptional activation complexes tether the yeast activator Met4 to DNA EMBO J 17 6327 6336 9799240
Dormer UH Westwater J McLaren NF Kent NA Mellor J 2000 Cadmium-inducible expression of the yeast GSH1 gene requires a functional sulfur-amino acid regulatory network J Biol Chem 275 32611 32616 10921921
Young ET Dombek KM Tachibana C Ideker T 2003 Multiple pathways are co-regulated by the protein kinase Snf1 and the transcription factors Adr1 and Cat8 J Biol Chem 278 26146 26158 12676948
Jungmann J Reins HA Lee J Romeo A Hassett R 1993 MAC1, a nuclear regulatory protein related to cu-dependent transcription factors is involved in cu/fe utilization and stress resistance in yeast EMBO J 12 5051 5056 8262047
Labbe S Zhu Z Thiele DJ 1997 Copper-specific transcriptional repression of yeast genes encoding critical components in the copper transport pathway J Biol Chem 272 15951 15958 9188496
Rep M Proft M Remize F Tamas M Serrano R 2001 The Saccharomyces cerevisiae Sko1p transcription factor mediates HOG pathway-dependent osmotic regulation of a set of genes encoding enzymes implicated in protection from oxidative damage Mol Microbiol 40 1067 1083 11401713
Coffman JA Rai R Loprete DM Cunningham T Svetlov V 1997 Cross regulation of four GATA factors that control nitrogen catabolic gene expression in Saccharomyces cerevisiae
J Bacteriol 179 3416 3429 9171383
Dodou E Treisman R 1997 The Saccharomyces cerevisiae MADS-Box transcription factor Rlm1 is a target for the Mpk1 mitogen-activated protein kinase pathway J Bacteriol 17 1848 1859
Cunningham TS Cooper TG 1993 The Saccharomyces cerevisiae DAL80 repressor protein binds to multiple copies of GATAA-containing sequences (URSGATA) J Bacteriol 175 5851 5861 8376332
Cunningham TS Cooper TG 1994 The UGA4 UASNTR site required for GLN3-dependent transcriptional activation also mediates DAL80-responsive regulation and DAL80 protein binding in Saccharomyces cerevisiae
J Bacteriol 176 4718 4725 8045902
Hongay C Jia N Bard M Winston F 2002 Mot3 is a transcriptional repressor of ergosterol biosynthetic genes and is required for normal vacuolar function in Saccharomyces cerevisiae
EMBO J 21 4114 4124 12145211
Linde JJT Steensma HY 2002 A microarray-assisted screen for potential hap1 and rox1 target genes in Saccharomyces cerevisiae
Yeast 19 825 840 12112237
Fernandes L Rodrigues-Pousada C Struhl K 1997 Yap, a novel family of eight bZIP proteins in Saccharomyces cerevisiae with distinct biological functions Mol Cel Biol 17 6982 6993
Pramila T Miles S GuhaThakurta D Jemiolo D Breeden LL 2002 Conserved homeodomain proteins interact with MADS box protein Mcm1 to restrict ecb-dependent transcription to the M/G1 phase of the cell cycle Genes Dev 16 3034 3045 12464633
Yona E Bangio H Ehrlich P Tepper SH Katcoff DJ 1995 The c-terminal domain of SIN1 in yeast interacts with a protein that binds the URS1 region of the yeast HO gene Mol Gen Genet 246 774 777 7898448
Novoseler M Hershkovits G Katcof DJ 2005 Functional domains of the yeast chromatin protein Sin1p/Spt2p can bind four-way junction and crossing DNA structures J Biol Chem 280 5169 5177 15563464
Chellappa R Kandasamy P Oh CS Jiang Y Vemula M 2001 The membrane proteins, Spt23p and Mga2p, play distinct roles in the activation of Saccharomyces cerevisiae OLE1 gene expression. Fatty acid-mediated regulation of Mga2p activity is independent of its proteolytic processing into a soluble transcription activator J Biol Chem 276 43548 43556 11557770
Rajewsky N Vergassola M Gaul U Siggia ED 2002 Computational detection of genomic cis regulatory modules, applied to body patterning in the early Drosophila embryo BMC Bioinformatics 3 30 12398796
Zavolan M Rajewsky N Socci ND Gaasterland T 2003 Smashing regulatory sites in DNA by human-mouse sequence comparisons IEEE Computer Society Bioinformatics Conference (CSB'03) 277 286
Sinha S van Nimwegen E Siggia ED 2003 A probabilistic method to detect regulatory modules Bioinformatics 19 i292 i301 12855472
Sinha S Siggia ED 2005 Sequence turnover and tandem repeats in cis-regulatory modules in drosophila Mol Biol Evol 22 874 885 15659554
National Human Genome Research Institute 2005 Saccharomyces Genome Database [database] Available: http://www.yeastgenome.org/ . Accessed 7 November 2005.
Rokas A Williams BL King N Carroll SB 2003 Genome-scale approaches to resolving incongruence in molecular phylogenies Nature 425 798 804 14574403
|
16477324
|
PMC1309704
|
CC BY
|
2021-01-05 09:18:23
|
no
|
PLoS Comput Biol. 2005 Dec 9; 1(7):e67
|
utf-8
|
PLoS Comput Biol
| 2,005 |
10.1371/journal.pcbi.0010067
|
oa_comm
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.